Latest / Elon Musk Podcast / Tesla's 2023 Investor Day: Insights on Expansion, Generation 3 Platform, and Capital Allocation

Transcript

Tesla's 2023 Investor Day: Insights on Expansion, Generation 3 Platform, and Capital Allocation · Elon Musk Podcast

• 0:00
  Quarters in Austin, Texas. We welcome you.
• 0:02
  Thank you for being here. And thank you for traveling in
• 0:05
  for those joining us virtually. Thank you for being a part of
• 0:08
  the day today. So, as we reflect back on the
• 0:12
  history of the company, there's been distinct phases of product
• 0:16
  advancement, Technology, Innovation and Rapid volume
• 0:20
  growth. The most recent of which has
• 0:23
  been the global expansion and localization of the model y
• 0:26
  program. And so, today, we want to talk
• 0:30
  about the future. We don't want to talk about this
• 0:33
  quarter or next quarter, we want to go further out into the
• 0:36
  future. And we've divided today's
• 0:39
  presentation into three parts. The first of which we're going
• 0:43
  to go macro. What does it take to convert
• 0:47
  Earth to sustainable, energy, generation and use The second we
• 0:53
  want to talk about Teslas contribution to that Global
• 0:57
  need. We're going to go function by
• 0:59
  function through the company and you'll meet our entire
• 1:03
  leadership team and we're going to get into the details of what
• 1:07
  those teams are doing as part of the broader goal.
• 1:11
  And then in the third part, we're going to bring it back up
• 1:15
  and talk about what this all means.
• 1:17
  For the company, as a whole So before we get started statements
• 1:23
  made in this presentation, are forward, looking statements that
• 1:26
  are subject to risks and uncertainties, more details can
• 1:30
  be found in our written materials.
• 1:33
  So with that, let's get started part 1, Elon Musk and Drew
• 1:38
  Bagley know. Alright, master plan part 3. so,
• 1:56
  as It's like was mentioning the thing that I think is we wanted
• 2:01
  to convey probably more importantly than anything else
• 2:05
  that we talked about here is that there is a clear path to a
• 2:10
  sustainable energy Earth. It's not it doesn't require
• 2:16
  Destroying natural habitats. It doesn't require us to be
• 2:26
  austere and stop using electricity and sort of being
• 2:29
  the cold or anything the story. And I think that this holds
• 2:34
  together quite well and will be actually publishing detailed
• 2:37
  white paper with all of our assumptions and calculations is
• 2:40
  that there is that there is a clear path to a fully
• 2:42
  sustainable Earth with abundance.
• 2:45
  In fact, you could support Art a civilization, much bigger than
• 2:49
  Earth than much, more than the the 8 billion.
• 2:51
  Humans could actually be supported sustainably on Earth
• 2:57
  and I'm just often shocked and surprised by how few people
• 3:02
  realize this. Most of the smart people.
• 3:08
  I know actually don't see a clear path.
• 3:12
  They think that there's there's not a path to a sustainable
• 3:18
  energy future or at least there's not one that is
• 3:22
  sustainable at our current population or that would have to
• 3:27
  resort to Extreme Measures. None of this is true.
• 3:31
  So we're going to walk through the calculations for how to
• 3:35
  create a sustainable energy. Civilization.
• 3:42
  And to set the stage today, our energy economy.
• 3:45
  It's let's be honest, it's dirty, and it's wasteful over
• 3:49
  80% of global energy, primary energy comes from fossil fuels
• 3:52
  and only one-third of that Global energy actually ends up,
• 3:55
  delivering useful work or heat. This is the problem statement
• 3:59
  but we're here to talk about the solution.
• 4:01
  Yeah it's like if for some of this I'm going to elaborate
• 4:05
  because there's a very wide range of technical expertise.
• 4:09
  He's out there from people who are like, you know, whatever
• 4:13
  level nine Wizards in the subject, to people who do not do
• 4:18
  engineering at all. So we're, if you have a gasoline
• 4:21
  car, you're you're converting less than a third of, and maybe
• 4:28
  only 25% of the energy in the gasoline is converted into
• 4:32
  motion. The rest is turn into waste
• 4:34
  heat. That does no doesn't do any good
• 4:36
  role and there's a lot of energy required even to get there or
• 4:39
  all other ground to refine the oil and to transport the
• 4:42
  gasoline to the gas station. So when you, when you look at
• 4:46
  all that for a typical gasoline, car is actually going to be
• 4:51
  using less than 20 percent fully considered of the energy from
• 4:56
  the oil actually goes into motion.
• 4:59
  So this is a when I see who, or when we see people doing
• 5:03
  calculations for what does it take to create a sustainable
• 5:05
  energy Earth? They assume that the same energy
• 5:08
  amount is required for an electric, or electric,
• 5:12
  electrified, civilization versus a combustion civilization.
• 5:16
  This is not true that because most of the energy of combustion
• 5:22
  is waste Heat. And even to get the fuel to
• 5:26
  combust in the first place and get it to the end use, there's a
• 5:28
  lot lost along the way. I mean, this is the primary
• 5:31
  energy consumption 165. PW h a year, PW H is a trillion
• 5:35
  terroir trillion kilowatt-hours. So it's a large amount of energy
• 5:39
  but the nice thing about electrified economy.
• 5:42
  There's a better way. We're going to talk about it is
• 5:45
  that through end-use efficiency and through efficiency along
• 5:49
  every step of the way actually the total energy use halves.
• 5:54
  So this is one of the most enabling aspects of electrifying
• 5:57
  everything is that the sustainable energy economy is
• 6:00
  that much easier to accomplish. It's actually half the problem
• 6:03
  statement of the fossil fuel economy.
• 6:05
  Yeah. And we're being conservative
• 6:07
  here so it could be better than half but we're trying to have
• 6:11
  assumptions that are reasonable and not overly optimistic.
• 6:14
  In fact slightly pessimistic so it's really better than half but
• 6:18
  just say for it's easy to make the argument that we need half
• 6:21
  as much energy with an electric. Me versus a combustion economy.
• 6:26
  Yep. So how the master plan works?
• 6:30
  You want to talk? Yeah.
• 6:33
  So the the thing that is needed in at very large scale, that is
• 6:40
  not currently present, is a vast amount of battery energy
• 6:44
  storage. Our rough calculations are that
• 6:48
  this is about 240 terawatt-hours over 240,000 gigawatt hours.
• 6:55
  This is a lot of batteries, but it is actually a very achievable
• 6:59
  Mount will go into details on that.
• 7:02
  So that's a combination of electric vehicles and stationary
• 7:06
  storage. So if you've got solar or wind,
• 7:08
  you've got to store the energy when the wind is not blowing,
• 7:11
  the sun is not shining and so we're assuming sort of an 8, to
• 7:15
  1 ratio of stored energy to power.
• 7:20
  So, 30 terawatt hours of power, that's it.
• 7:22
  So, there are three terawatts of power our actual Of expenditure
• 7:29
  calculation for manufacturing investment is more like Six
• 7:34
  Trillion, but we, you know, we should made it higher to make it
• 7:39
  ten trillion and this is across mining refining, you know,
• 7:43
  battery factories, recycling, vehicle factories, all the
• 7:45
  things that we're going to talk about needing to invest in to
• 7:48
  build this sustainable energy economy.
• 7:50
  Yeah. Now, if you look at the total
• 7:53
  World economy, it's just under a hundred trillion.
• 7:56
  So if this was spread out say over, Ten years, it would be one
• 8:00
  percent of the global economy. Over 20 years it would be half a
• 8:04
  percent are adorable economy. So this is yeah, not a big
• 8:09
  number relative to the global economy.
• 8:13
  As Drew mentioned, you need about half as much energy with
• 8:16
  an electric Connie, versus a combustion economy and in terms
• 8:20
  of wind and solar, how much land would be used.
• 8:23
  It's less than point two percent of the land area of Earth.
• 8:29
  Like generally, people don't realize quite how much energy is
• 8:32
  reaching us from the sun. It's roughly a giga watt per
• 8:36
  square kilometer and sun doesn't shine all the time.
• 8:41
  But it's if you multiply that by say, For it to get, the
• 8:47
  continuous power, 4 or 5, then that that gives you the land
• 8:52
  area of solar and you can put wind and solar often in the same
• 8:55
  place. So a lot of places that
• 8:57
  currently have wind, you could have solar there and you double
• 9:00
  your energy, you can also put wind offshore.
• 9:02
  It doesn't even need to be on land.
• 9:03
  So when does even more flexible could put solar offshore to.
• 9:07
  Yep. So Earth is 70% water.
• 9:11
  What is that? Whether the pretty really
• 9:14
  remarkably small amount of Earth's land area.
• 9:17
  We can go fully sustainable. Yeah.
• 9:22
  And and and from a to the resources and raw materials
• 9:25
  exist to support this transition, will go through
• 9:28
  that. In detail about, we do not see
• 9:30
  any insurmountable resource challenges at all.
• 9:32
  In fact, in the end we should be mining less or to accomplish
• 9:36
  this economy than we currently do with the fossil fuel economy
• 9:38
  and we're going to talk through that.
• 9:40
  Yeah, just emphasize that again. The electrified economy will
• 9:44
  require less mining than the current economy does.
• 9:48
  Yes, less not more. Okay.
• 9:51
  So this is the plan and now we'll get into a little more of
• 9:54
  the details of the plan. Basically five areas of work,
• 9:59
  first row will power, the existing grid second switch to
• 10:04
  the to electric vehicles. Third, switch homes businesses
• 10:07
  and Industry heating to heat pumps.
• 10:10
  Fourth high temp heat delivery and storage for high temp,
• 10:15
  industrial and chemical processes.
• 10:17
  And a little bit of green hydrogen in there for chemical
• 10:20
  processes that need hydrogen. And finally, sustainable
• 10:24
  sustainable, fuel planes and boats.
• 10:25
  These are the five areas and we're going to go into detail on
• 10:27
  all of them. Yeah, I mean, my personal
• 10:29
  opinion is that as we improve the energy density of batteries
• 10:33
  you'll see all Transportation go fully electric with the
• 10:39
  exception of rockets, there's awkward.
• 10:43
  But could you can make the fuel with CO2 and water so you can
• 10:48
  make methane with CO2 and water. So in fact that With just
• 10:52
  electricity's. Yes, exactly.
• 10:53
  So, so in fact on Mars, we hopefully get there at some
• 10:58
  point, the atmosphere CO2 and there's water ice thread Mars so
• 11:03
  you can take the CO2 and H2O and turn that into CH4 which is
• 11:08
  methane and oxygen. So ultimately, even Rockets can
• 11:12
  be electrified. So first repowering, the
• 11:17
  existing grid with Renewables and this is going to be a
• 11:19
  consistent theme, you'll see our estimates for the number of
• 11:22
  terawatt-hours terawatts and trillions of investment at the
• 11:25
  bottom of the page. You know, this is already
• 11:28
  actively occurring in front of us. 60% of the generation added
• 11:31
  to the US grid was solar in 2022 and actually on a year-on-year
• 11:36
  basis solar deployment is growing 50 percent year-on-year
• 11:40
  as a 2022. So this is a this is a serious
• 11:45
  Upswing. And if we continue this trend,
• 11:47
  this is going to be behind us before we even know it.
• 11:50
  Yeah. Second switching to electric
• 11:53
  vehicles again. 21% reduction in fossil fuel use by doing this
• 11:57
  alone. Obviously, Tesla's heavily
• 12:00
  engaged in this activity as along, with many others overall,
• 12:04
  EV production, grew 59 percent year on year in 2022 and V's hit
• 12:08
  an amazing 10% market share. I mean, it's an awesome
• 12:12
  Milestone. I'm super excited to see that.
• 12:15
  and yeah, this is Obviously happening.
• 12:21
  Very rapidly and I mean, I think really old cars.
• 12:26
  Will go to fully electric and autonomous and so writing a non
• 12:34
  autonomous gasoline car is going to be analogous riding a horse
• 12:38
  and using a flip phone. That's basically the going to be
• 12:41
  the situation and we actually took a somewhat conservative
• 12:44
  assumption here in terms of how many batteries are required
• 12:46
  because the more the fleet is autonomous.
• 12:48
  The Fewer of the smaller the fleet needs to be just from
• 12:52
  utility rate basis. So we're not accounting for all
• 12:55
  of those benefits really much of those benefits at all in this
• 12:57
  number and what does this Fleet look like, you know, just rough
• 13:02
  view from our perspective, of course, we could be wrong but
• 13:05
  you know, you can see the sort of breakdown of the fleet by
• 13:08
  millions of vehicles. You know, our goal is to do 20
• 13:12
  million electric vehicles a year.
• 13:14
  Yeah, fewer Vehicles will be needed at least passenger
• 13:18
  vehicles. With autonomy.
• 13:21
  So Debate as to what that number is.
• 13:25
  But it's some number less than the number of vehicles needed
• 13:29
  today. There's roughly 2 billion cars
• 13:30
  and trucks in operation in the world today.
• 13:35
  Yeah. So what we show here is actually
• 13:36
  I think only one point four million or so.
• 13:40
  So we represent one point four billion.
• 13:42
  I mean are so so a smaller Fleet and you know, the numbers are
• 13:46
  here in this presentation are around 85 million Vehicles, a
• 13:49
  year produced, just to give you a sense of how we're thinking
• 13:51
  about this. Again, we're going to put all
• 13:52
  these assumptions. Up online and you know,
• 13:55
  encourage people's thoughts. Yeah.
• 13:58
  So yeah. So we're basically heading
• 14:01
  directly towards an electric or autonomous future exciting.
• 14:06
  Yeah and one of the reasons why V's are so enabling is this
• 14:10
  end-use efficiency Point Tesla Model 3, it's four times more
• 14:15
  efficient from the well to wheel than a Toyota Corolla and that's
• 14:20
  all about the efficiency of getting the electricity to the
• 14:23
  into the car, in a sustainable and energy economy and that how
• 14:26
  efficient the cars in transferring that stored energy.
• 14:30
  G2 motion on the road when compared to the engine in the
• 14:33
  Toyota Corolla, and all the, you know, extraction refining.
• 14:37
  Transmission distribution of the gasoline to the Toyota Corolla.
• 14:40
  And just for this is a fun reference model 3, can drive
• 14:44
  over a mile on the energy. It takes to boil a pot of water
• 14:46
  for pasta, and then it can drive another mile on the energy.
• 14:50
  It took to cook the pasta, And that pasta is one pound them.
• 14:54
  All three is like 4,000 pounds just to give you a sense of just
• 14:58
  like, what like it really doesn't use a lot of energy to
• 15:01
  move the model, 3 that 4,000-pound object down the road
• 15:05
  also, heat is lot more energy than motion.
• 15:07
  Yeah. Yeah.
• 15:08
  If you but people, you know, you just boil a pot of Asia.
• 15:10
  Let me think about it. Yeah, it's just interesting.
• 15:13
  How efficient these cars are next switching to heat, pumps
• 15:17
  and homes businesses and Industry.
• 15:20
  You know, right now he pumps meet 10% of building heating
• 15:23
  needs. Install rates.
• 15:24
  Growing ten percent year-on-year.
• 15:25
  It really needs to accelerate heat pumps can serve you.
• 15:28
  No heat applications up to 200 C and businesses and Industry and
• 15:32
  from an investment perspective. As you can see, on this page is
• 15:35
  actually the lowest hanging fruit in terms of displacing
• 15:38
  fossil fuels. You might be saying, well, what
• 15:40
  exactly is a heat pump? So, He pumps don't move heat,
• 15:47
  don't create heat, they move heat when you think about like
• 15:51
  the natural gas furnace in your house like it's it is generating
• 15:54
  heat itself. But what the heat pump is doing
• 15:56
  is actually moving heat from outside of your house into your
• 15:58
  house, they're just an air conditioner or a refrigerator in
• 16:01
  Reverse. So we're surrounded by heat
• 16:03
  pumps. There's like you know they're
• 16:05
  all over this Factory there in your house and all this really
• 16:08
  is is about bringing them to displace.
• 16:11
  All the fossil fuel Heating in all the homes businesses and in
• 16:14
  the industry. That we can.
• 16:15
  And from Deus efficiency perspective is a 33 times
• 16:19
  reduction, in the total energy required to heat these
• 16:21
  buildings. So, I real obvious thing to do.
• 16:25
  Yeah, he pumps. There are no Cars, ya know it,
• 16:29
  as default. And at some point we might make
• 16:32
  heat pump for home. So yeah, maybe maybe maybe next
• 16:38
  a little bit more detail on electrifying, high temp, sort of
• 16:42
  industrial chemical processes so over 50% of industrial heat is
• 16:46
  greater than 400 see where you know cement steel fertilizers
• 16:49
  chemicals Plastics Metals refining all need like 1500 C so
• 16:53
  we need to select Lucien here. Ultimately, it's purpose-built
• 16:56
  equipment that enables electrification, you know,
• 16:59
  Carbon graphite is stable up to 2,800, 2,800.
• 17:02
  See, there's other options in the 1500, see range, like
• 17:05
  silicon carbide, other materials.
• 17:08
  So the idea here is you create and store heat, when removal
• 17:12
  power is available. If this is sustainable energy
• 17:14
  economy, you know renewable energy is intermittent peak of
• 17:17
  the day, you've got more generation than you need, you
• 17:20
  make a bunch of heat then and then you transfer that heat into
• 17:23
  the end. The process 24 hours a day using
• 17:25
  the stored heat, you created when the Sun or the wind was
• 17:28
  blowing. That's the concept here.
• 17:30
  And then on the hydrogen side, we also need green.
• 17:32
  Hydrogen to decarbonize, metals and chemical refining processes.
• 17:35
  This is things like ammonia making steel.
• 17:39
  You know, there's roughly 120 million, tons of hydrogen Source
• 17:42
  from fossil fuel today, to do these things.
• 17:45
  And hydrogen can also directly replace coal, which is currently
• 17:48
  used in a ton of steel production through a process
• 17:51
  called directly reducing iron, you can replace It's blast
• 17:54
  furnaces with a hydrogen reducing direct reduced iron
• 17:58
  furnace. And this is the way to
• 18:00
  illuminate fossil fuels, from these aspects of the economy and
• 18:04
  the CO2 associated with them. Yeah.
• 18:05
  So I mean, some of this, this like room to disagree, but some
• 18:12
  amount of hydrogen is needed for industrial processes.
• 18:14
  My personal opinion is that hydrogen will not be used
• 18:17
  meaningfully and transport but and shouldn't be its.
• 18:22
  If you're going to use a chemical Fuel, you should use
• 18:25
  CH4, not H2, but it's not unless it is needed for industrial
• 18:30
  processes and can be produced just by splitting water.
• 18:34
  Essentially. I mean, some of that's been done
• 18:36
  for decades and decades, this is not, you know, rocket
• 18:39
  technology. Yeah.
• 18:43
  And lastly, a small part of the pie but a necessary part of the
• 18:46
  pie is sustainably, fueling planes and boats shipping
• 18:50
  accounts for 3% of global CO2. It's right for electric.
• 18:53
  Vacation, even with lithium iron, phosphate long-haul ships,
• 18:57
  can be fully battery powered, so that's a great opportunity to
• 19:01
  Electrify energy. Density is a little bit harder
• 19:04
  for planes, but short-haul is doable today with some
• 19:06
  improvements, will get long-haul underway but even even in the
• 19:09
  meantime we can leverage sustainable Aviation fuels
• 19:12
  producing stored using access renewable electricity.
• 19:14
  There's a lot of work going on in this space.
• 19:17
  And it's yeah, I mean it to really get long range aircraft.
• 19:23
  And long-range shipping to use lithium-ion, you need to
• 19:30
  redesign the ship and not just or the plane.
• 19:33
  And the plane to take advantage of the fact that it is a new
• 19:39
  source of energy as it's a different architecture.
• 19:44
  So, just like with an electric car, you wouldn't just, you
• 19:48
  know, take a gasoline car and stick a battery in it.
• 19:50
  That's very suboptimal. It's much more efficient to have
• 19:55
  the battery. Be the structure of the car and
• 20:02
  make it as make it Mass, efficient and optimized for
• 20:05
  veterans or veterans the say. If that's done with the aircraft
• 20:08
  that you, get long-range aircraft at around with with
• 20:12
  sales at around 450 Watt, hours per kilogram, which you can buy
• 20:16
  it right now, actually the expensive, but I think that
• 20:19
  price will come down So when we stack up all of these efforts,
• 20:24
  we end up with the numbers. We shared at the beginning of
• 20:26
  the presentation, 30 TW 240, terawatt-hours, 10 trillion
• 20:30
  dollars. And your you may be saying like
• 20:33
  I need some context is this feasible.
• 20:37
  Spoiler alert, it's entirely feasible, just looking at it
• 20:41
  from a growth rate of growth rate perspective.
• 20:44
  How much do we need to grow the deployment of these
• 20:46
  Technologies. We're talking about only a 3 x 2
• 20:49
  growth rate and solar and wind deployment solar is already
• 20:53
  growing at a Breakneck Pace. As is wind this Gap is going to
• 20:57
  be close really quickly. When we look at the electric
• 21:00
  vehicles they have to grow 11x. Well they grew 60 percent year
• 21:04
  on year last year that That growth rate is also going to
• 21:07
  close pretty, pretty darn. That Gap is going to close
• 21:09
  pretty quickly as well. And lastly storage, you know
• 21:14
  Tesla's energy storage business has grown at 65% kakkar.
• 21:17
  Since 2016, the global you know, energy storage businesses is
• 21:22
  accelerating Pace as well. I mean all these gaps are going
• 21:25
  to close, especially as as this momentum of the transition to
• 21:28
  sustainable energy, accelerates, and of course, our goal on this
• 21:33
  page is 20 million TVs per year and one Terawatt hour of
• 21:37
  stationary storage per year, basically, as soon as we can,
• 21:43
  and then, you know what about this investment?
• 21:46
  How do I have a reference point on this investment, you know,
• 21:48
  Elon mentioned? It's 10% of, you know, one
• 21:52
  year's world GDP. Another way of thinking about it
• 21:55
  is, how does it compare to what we're investing?
• 21:57
  Like what we invested last year in the fossil fuel
• 22:00
  infrastructure and, and it's 60% of that investment.
• 22:03
  So, actually building the All energy economy is less than
• 22:08
  extending the fossil fuel economy from year over year
• 22:12
  investment basis. So very doable when we look
• 22:16
  towards, does this fit on the planet?
• 22:19
  Absolutely less than 0.2 percent of land as a reference point,
• 22:25
  the total land area intensively, farmed today is 12 and a half
• 22:30
  percent of all land. So I mean, you drive around, you
• 22:33
  see some Farms but you don't see them everywhere.
• 22:36
  This is this is an order of magnitude more than order of
• 22:38
  magnitude difference between farming and what we're talking
• 22:40
  about for sustainable energy land, and it doesn't need to
• 22:44
  displace Farmland. Of course.
• 22:46
  Yeah. You know, forests or jungle, or
• 22:49
  any kind of ecological, preserve.
• 22:53
  It can be used and very sparsely populated desert regions.
• 22:57
  Barren areas areas that are just not really fit for development
• 23:01
  or otherwise you so yeah, I mean, point two percent can fit
• 23:04
  into a lot of places. Yeah.
• 23:06
  It's there's essentially no meaningful ecological impact.
• 23:11
  In fact, transition to a sustainable energy Connery.
• 23:14
  Economy would result in a substantial reduction in current
• 23:18
  ecological impact. It's a great way to put it.
• 23:20
  Yeah. And what about on the mineral
• 23:23
  extraction side? So this is a cartoon that sort
• 23:26
  of gives you a sense for all the or and the like extracted
• 23:30
  minerals that are coming out of the Earth every year, it's about
• 23:34
  68 gigaton. Tons.
• 23:37
  So each truck is a good time. What does this look like when
• 23:40
  we're in a sustainable energy economy?
• 23:42
  Looks like that. Fossil fuel extraction
• 23:46
  disappears. We replace it with the materials
• 23:48
  required to fulfill the sustainable energy economy
• 23:51
  actually reduces. Now it's not to say that we
• 23:54
  don't need to continue to explore, you know, bring on
• 23:57
  mining and refining for the sort of specific materials for the
• 24:01
  sustainable energy. Economy we do.
• 24:03
  But the investment in Mass flows are all very achievable just
• 24:05
  looking at what is already happening on the planet.
• 24:07
  Like this is nothing out of scale of what has been done and
• 24:10
  is already being done. And then we calculate it on a
• 24:15
  sort of, you know, element by element basis.
• 24:18
  The resources are there to support the transition.
• 24:21
  You know, this is cumulative demand to move in the
• 24:24
  sustainable energy economy Direction until 2050 relative to
• 24:27
  USGS resources today. You know, we're not breaking the
• 24:31
  Resource Bank for any of these materials and then when we look
• 24:35
  at what really happens as we move forward, History teaches,
• 24:39
  the more we look, the more we find.
• 24:41
  What people think happens is, oh, there's this many resources
• 24:44
  next year. There's going to be less because
• 24:45
  we're going to extract them. What actually happens is as we
• 24:50
  extract resources, we find more and you can see on the right
• 24:53
  what is actually occurred with the key materials to the
• 24:56
  sustainable energy economy since 2000.
• 24:58
  As a sustainable energy economy has been growing and test has
• 25:00
  been growing and all the industries around us have been
• 25:03
  growing. The actual resource availability
• 25:05
  has increased not decreased. Yes.
• 25:09
  There's this little seems to be quite a bit of confusion about
• 25:12
  the lithium. Lithium is extremely.
• 25:14
  Common is one of the most common elements on Earth.
• 25:17
  There is, there's no country that has Monopoly on lithium or
• 25:21
  even close to it. There's there's enough lithium
• 25:26
  or in the United States to Electrify all of Earth.
• 25:29
  If the United States was the only place producing with you,
• 25:32
  there's enough domestic material to Electrify Earth.
• 25:37
  It's very common. Me the, the limiting limiting
• 25:41
  factor is the refining of the lithium into battery grade,
• 25:46
  lithium hydroxide or lithium carbonate.
• 25:48
  That's the actual limiting factor and the same is true for
• 25:50
  these other materials. And, and this is again, these
• 25:53
  are not like crazy Technologies. It's just the Investments need
• 25:55
  to be made and the Investments. They're not gigantic, they just
• 25:58
  need to happen, right Nichols. Maybe the thought of them all
• 26:03
  the trickiest one to to solve, but as we showed with the graph,
• 26:07
  there may be need like 30 percent of the world's.
• 26:09
  No, nickel reserves. So, and the nickel reserves have
• 26:11
  actually groans. Yes, thousand.
• 26:13
  There is morning. Exactly.
• 26:15
  So and you only need nickel for basically aircraft long-range
• 26:21
  boats and very long range cars or trucks.
• 26:25
  But the vast majority of the heavy lifting for
• 26:29
  electrification will be iron-based cells.
• 26:34
  And Earth is iron, is actually literally the most common
• 26:37
  elements on Earth. So, little trivia point if you
• 26:42
  say like, what is Earth made of by mass, it is most, it is made
• 26:47
  of iron more than anything else and second oxygen, and then
• 26:53
  everything else after that. So, basically, we're muddy rust.
• 26:57
  Ball is what Earth is so an iron cathode is sort of, definitely
• 27:03
  not going to run out of iron. There's so much.
• 27:04
  Iron is insane. So That's yeah yeah and
• 27:12
  ultimately you know, this resource extraction, we go
• 27:17
  through this effort, we build these batteries and then we
• 27:20
  recycle these batteries. So, ultimately, we're doing this
• 27:24
  to build this to Sanibel energy economy, but the maintenance
• 27:26
  amount of of, or that we require is really an order of magnitude
• 27:30
  or more less because of recycling.
• 27:34
  So, in the end, A sustainable energy economy is within our
• 27:39
  reach and we should accelerate it.
• 27:41
  Yeah. I mean so that this is really
• 27:44
  the main message of today and I really want today to be not just
• 27:49
  about tells the investors who own stock, but really anyone who
• 27:53
  is an investor in Earth and the we're trying to convey is a
• 27:59
  message of Hope and optimism and hope optimism.
• 28:04
  That is based on an actual physics.
• 28:06
  And And and real calculations not it's not wishful, thinking
• 28:11
  Earth can and will move to a sustainable energy economy and
• 28:16
  will do. So in your lifetime, Thank you,
• 28:20
  yes, thank you. And I just want to welcome up
• 28:29
  Lars and Franz to the stage. Hi.
• 28:41
  Hi. I am Franz, I lead design at
• 28:45
  Tesla and I'm Lars. I've been doing vehicles with
• 28:49
  France for almost 13 years now. So, I joined Tesla and 2008 to
• 28:56
  vertically integrate design into the company didn't exist before.
• 29:00
  It was pretty small team and was tasked with designing most
• 29:03
  beautiful, Innovative and well-engineered vehicles on the
• 29:06
  planet. No small task in 2008.
• 29:09
  There's not a lot going on on the TV sector then since then we
• 29:13
  focused on constant Improvement and cost efficiency Innovation
• 29:17
  things that you'll continue to hear about today while
• 29:20
  continuing to design the most desirable cars.
• 29:25
  Today we produce cars differently than we did 10 years
• 29:27
  ago, but the end result is always an exciting futuristic
• 29:31
  and desirable set of vehicles back.
• 29:33
  Then we only had a handful of designers and Engineers like
• 29:36
  myself but we had a great vision to radically change
• 29:38
  transportation. So back in 2008, where were
• 29:43
  designing Model S, we didn't have a factory.
• 29:47
  In fact, we had a really small engineering team and a tiny
• 29:50
  design team. But that loud designed to lead
• 29:56
  all the conversations it led us innovate forward-thinking ideas
• 30:00
  like how do you fit seven people into a sedan or how do you make
• 30:04
  door handles disappear into the into the doors or putting a huge
• 30:09
  touchscreen into the center of the vehicle?
• 30:11
  Something that had never been done before.
• 30:15
  And then we want another chain Car of the Year.
• 30:16
  Yes. And we won more to trying to
• 30:17
  Carter the year in 2013, our first, our first you know, great
• 30:21
  award first car pretty good, pretty good start kind of a home
• 30:25
  run I think but that that whole process resulted in a linear
• 30:30
  process that you see on the screen we designed first then we
• 30:33
  engineered and then we figure out how to manufacture it.
• 30:36
  Yeah. I mean the think that's really
• 30:37
  important when we were designing the car together, we didn't even
• 30:40
  know where we're going to build it.
• 30:41
  And so we came up with me and no Lars wondered if that's true.
• 30:46
  We didn't say once we got Fremont, we were very fortunate
• 30:48
  and we figured out the manufacturing Solutions, sort of
• 30:51
  like we were flying a plane and putting the wings and building
• 30:53
  the engine at the same time. So we knew we had to do better.
• 30:57
  Yeah, we knew he had to be better in order to scale.
• 31:00
  And as part of the master plan that you've read model 3 needed,
• 31:04
  to be a smaller more efficient and more affordable version of
• 31:07
  Model S, but had to be equally, great heads up all the things
• 31:11
  that people loved in their Model S or model acts.
• 31:14
  And And what be much more affordable.
• 31:17
  And so we approached the process a little bit differently than
• 31:20
  the first time around. Now, we had teams that we all
• 31:24
  work together. So we're able to combine design
• 31:26
  engineering and Manufacturing process all at the same time.
• 31:31
  But somewhere along the way, we change the manufacturing process
• 31:36
  to be fully automated. And so, we leaned into this
• 31:39
  whole new way of manufacturing, a car, but we had already
• 31:42
  engineered it. So things didn't quite go as
• 31:45
  well as planned. It was an amazing product but it
• 31:49
  landed Us in production. Hell, many of us who lived
• 31:52
  through that. Carry those Battle Scars.
• 31:55
  It was a great idea but it wasn't the right timing.
• 31:58
  Like Franz said automating something that we designed to be
• 32:00
  built man. Julie is super hard and we have
• 32:03
  many, many failed examples of that at the Fremont Factory that
• 32:06
  we ripped out, but some of them are eventually.
• 32:08
  So work, this is one, I actually worked on with a small team of
• 32:11
  Engineers. It's still running today and
• 32:13
  some of the engineers that came by said we couldn't do it or no
• 32:16
  longer with the company, but it's running and it's running
• 32:18
  faster than ever. So, we kind of self-imposed
• 32:23
  constraints on the design when we were doing it to be built
• 32:27
  manually and we really didn't think about it.
• 32:29
  But despite all that model 3 is the best.
• 32:31
  Selling EV ever. Yeah, and model Y which is
• 32:35
  derived from model 3 is about to pass that.
• 32:40
  But we knew we had to improve the process further and with
• 32:43
  cyber chalk. We designed a vehicle around a
• 32:47
  vision that actually started with the manufacturing process.
• 32:50
  And in this case, the materials dictated, the design forming
• 32:55
  full heart, stainless steel isn't rocket science, but it
• 32:58
  sure isn't easy and it limited the way we could do it.
• 33:01
  Yeah. Absolutely.
• 33:02
  It really forces us to think about designing something in a
• 33:06
  way that you couldn't normally stamp pad annals.
• 33:09
  You could inform them in a traditional way.
• 33:10
  So you ended up with very linear bending processes that are just
• 33:16
  not in automotive, kind of language of manufacturing today,
• 33:20
  but it actually created a very efficient and process.
• 33:23
  And one of the most dynamic designs ever, I believe it's
• 33:27
  definitely something that's going to change the road
• 33:29
  landscape. Hopefully, you guys saw it down
• 33:31
  there and you experienced it. It's definitely real.
• 33:34
  Those are real trucks. We're on our way to build them,
• 33:36
  but what that stainless Opportunity did for us.
• 33:40
  It is, let us rethink the factory footprint.
• 33:42
  We don't Stamp Those. That's a huge part of it.
• 33:44
  We don't even paint them. So, our footprint got smaller,
• 33:47
  and we started to think about innovative ways to take those
• 33:50
  constraints and make great products, but that constraint
• 33:54
  didn't really change. The end result of the charkh,
• 33:57
  it's a super Dynamic truck and it has all the functionality you
• 34:00
  would expect out of any of the other competitive charts and the
• 34:04
  best thing about it. It's coming this year so,
• 34:14
  ideally After all that we would design engineer manufacturer and
• 34:19
  plan for automation happening together.
• 34:21
  It gives us the opportunity to question requirements.
• 34:24
  This is something that is fundamentally only available at
• 34:27
  Tesla. The places I used to work in the
• 34:29
  top manufacturing companies in the world.
• 34:31
  They don't sit together. Yeah, you're right.
• 34:33
  One team know where I know has all these teams together
• 34:36
  thinking about these processes from the very beginning.
• 34:39
  In fact, all of those engineering teams, manufacturing
• 34:42
  design automation, they're all in one work.
• 34:44
  They all report to one person. We can't Point fingers at each
• 34:47
  other. So we have to solve them
• 34:48
  together which is the best way to innovate the traditional way
• 34:52
  of making a vehicle. Is this you stamp it?
• 34:54
  You do build a body and white, you paint it and you do final
• 34:56
  assembly. And what's interesting is these
• 34:59
  shops are dictated by the the organizational structures that
• 35:03
  exist and their dictated by the boundaries that exist in the
• 35:07
  factories that are laid out. If something goes wrong in final
• 35:10
  assembly, you block the whole line and you end up with
• 35:12
  buffering in between this is at the tail end.
• 35:17
  Of its manufacturing. Optimization Henry, Ford, first
• 35:20
  invented this assembly line 1922, it's been 100 years and
• 35:24
  it's really hard to make a change after 100 years.
• 35:28
  And when you watch it happen it's actually really silly to
• 35:31
  guy like me, you take all these stamp panels, you put them
• 35:34
  together, then you put them in a framing station, you build a
• 35:38
  body? That looks something like a car?
• 35:40
  You put the doors on and then you paint them.
• 35:45
  Once you get the color, you take the doors off and then you start
• 35:49
  putting the interior inside the car.
• 35:52
  It comes in through the openings that already exists.
• 35:54
  I wish it went in like this big piece like this, but there's
• 35:56
  actually people coming in and out of the car.
• 35:59
  There's awkward movements. Then we lift the car up.
• 36:02
  We put stuff from underneath it, we put it down, then we put the
• 36:05
  seats in the car. And finally, we close it all out
• 36:08
  with glass and we bring those doors and went away for a trip
• 36:11
  and we put them back in the car. Most of the time we're doing
• 36:16
  this with a big giant car and moving it and doing really
• 36:19
  nothing to it at all. What's funny though?
• 36:21
  In this kind of whole process is that just recently Toyota just
• 36:24
  call this and Engineering work of art through the model.
• 36:31
  Why that that was humbling. But at Tesla, it's not good
• 36:36
  enough. If we're going to scale the way
• 36:38
  we want to do, we have to rethink manufacturing again.
• 36:41
  It's part of the master plan. We have to make another step.
• 36:44
  Engine cost. We started this on model.
• 36:47
  Why, when we made these huge Giga castings and we deleted,
• 36:50
  hundreds of parts. We simplified assembly with the
• 36:53
  model eyes. Structural battery, where we
• 36:56
  decided the floor should be a part of the car.
• 36:58
  The battery is the floor. We put the seats in the interior
• 37:02
  on the battery and we bring it up through a big open hole and
• 37:05
  we assemble it in. This allowed us to do things in
• 37:08
  parallel, fully rethinking the process and reducing the final
• 37:11
  assembly line by about 10%. And we thought maybe we could do
• 37:15
  this other places. Yeah, I mean in a way the the
• 37:20
  constraints become part of the solution rather than a problem.
• 37:24
  So when you think about what I'm trying to say, I really want to
• 37:27
  hammer this home. When you have a car that's about
• 37:29
  5 meters long and you have people working around it like we
• 37:32
  did in model 3 and we change that to this process where we
• 37:36
  take different parts of the car and we can do more at the same
• 37:39
  time like we did with the model why structural battery pack,
• 37:42
  what you see here is US doing that on the front part of the
• 37:45
  vehicle or the rear part of the vehicle.
• 37:48
  That means we can get more people working on the car or
• 37:50
  robots, working on it at the same time.
• 37:54
  Means we have better operator density less time doing nothing.
• 37:58
  I call that space-time efficiency.
• 37:59
  It has nothing to do with quantum mechanics.
• 38:01
  We can have that conversation later but we get 44%.
• 38:05
  More operator density, which means more work, less time
• 38:08
  walking back to the station, 30 percent Improvement, and
• 38:11
  space-time efficiency. And because we're not building
• 38:15
  it in and out of the car with those slow movements of those
• 38:17
  robots, I showed you from production how, when we go to
• 38:20
  automate it, it's going to be a lot easier.
• 38:24
  In the end that will probably look something like this, where
• 38:26
  we balance parallel and series Manufacturing in a way where we
• 38:30
  only do things that are necessary with a much shorter.
• 38:33
  Final line blocking a lot less of the entire rest of the
• 38:36
  factory so we can optimize material flow using the best
• 38:39
  practices. And what that means it's going
• 38:41
  to look something like this where we build all the sides of
• 38:44
  the cars independently. We only paint what we need to.
• 38:48
  And then we assemble the parts of the car once.
• 38:52
  And only once we put them where they They need to go the
• 38:55
  Interiors tat attached from a bottom-up or top-down strategy.
• 38:59
  So there's more access for those robots and people we aren't
• 39:02
  moving heavy objects around and doing nothing to it and it means
• 39:06
  we're doing more work on the car more of the time.
• 39:09
  Then when we take it, all of these tested subassemblies and
• 39:13
  we put them together, we finally assemble the car only one time
• 39:17
  putting the sides on with all of their parts to a front and rear
• 39:20
  that was already assembled carrying the floor in with the
• 39:22
  seats. And finally, Sing it out with
• 39:25
  the doors one time just like cyber truck.
• 39:28
  So in the end you get the same car.
• 39:32
  But it's not going to be a model while.
• 39:34
  Yeah, this is not, this is a model Wyatt for illustration.
• 39:37
  Not the next-gen vehicle in the end.
• 39:41
  What does that mean to increase the scale and Adoption of
• 39:44
  electric vehicles on the orders of magnitude that we just showed
• 39:46
  you? We have to make constraints part
• 39:48
  of the solution. It leads us to greater than 40
• 39:51
  percent reduction in footprint which means we can build
• 39:54
  factories faster with less capex and more output per unit.
• 40:00
  Faster, less capex, more output per unit dollar Zacks.
• 40:05
  Going to go into more details on this later, but it also means
• 40:08
  through this Innovation and some of what my other engineering
• 40:11
  colleagues are going to talk to you about the future will reduce
• 40:14
  cost as much as 50%. This is the two-for-one concept.
• 40:17
  You hear me an Elon talking about on earnings calls.
• 40:22
  Yeah. So I think our track record
• 40:23
  approves that we can deliver the best cars and we deliver the
• 40:27
  best cars in spite of because of He's constraint and I'd love to
• 40:33
  really show you what I mean and unveil the next gen car.
• 40:37
  But you're gonna have to trust me on that until later date.
• 40:41
  Just and I promise we will always be delivering exciting
• 40:44
  compelling and desirable Vehicles like as we always have
• 40:47
  we ever not always do. But what else is Tesla known for
• 40:51
  Speed? So with that, I want to bring on
• 40:54
  Colin, who's in charge of making our cars go fast.
• 41:02
  Thank you Lars. Thank you Franz.
• 41:05
  My name is Colin Campbell and I have the real privilege of
• 41:08
  leading powertrain engineering here at Tesla.
• 41:11
  We make the fastest cars that you can buy for the money
• 41:14
  whether they're electric or gas, and the model S plaid that we're
• 41:18
  looking at here. It has more than 1,000
• 41:19
  horsepower and pound per pound. The motors in that car or as
• 41:23
  powerful as jet engines are, cars are super fun people,
• 41:27
  absolutely love driving in them. And the other thing that you
• 41:30
  probably all know about our powertrains is that they're
• 41:32
  efficient Arcos. Go 25 to 30 percent further than
• 41:37
  other EVS, in our class, for the same amount of efficiency.
• 41:41
  But at Tesla efficiency means more than just reducing how much
• 41:45
  energy the cars use. It's about how we develop, how
• 41:48
  we manufacture, how we refine, and how we scale the powertrain.
• 41:54
  Now the model 3 and Y, powertrain is a great example of
• 41:57
  this broader meaning of efficiency.
• 41:59
  So since we launched it back in 2017 we continuously improve
• 42:03
  that powertrain and the factory that builds it.
• 42:07
  So the Drive Unit, the engine of the car is 20% lighter for the
• 42:10
  same power, we use 25% less heavy rare Earths than when we
• 42:16
  started. And the powertrain the
• 42:18
  powertrain Factory which is behind me today is 75 percent
• 42:22
  smaller, and 65 percent cheaper than the one that we originally
• 42:25
  built. And what I really want to
• 42:28
  emphasize is that we did all of this without compromise.
• 42:32
  Our cars are just as powerful. They go.
• 42:34
  Just as far they cost the same or less and the factories have
• 42:38
  the same output. So, how did we do that?
• 42:45
  We did it by designing the entire Vehicle and the entire
• 42:48
  Factory together as one company. And this sets Tesla apart, we
• 42:54
  have small and highly capable teams and to make a critical
• 42:58
  decision. We can have the battery cell
• 42:59
  chemists, the mechanical engineers, the manufacturing
• 43:02
  Engineers, the supply chain team, the automation designers,
• 43:06
  the software programmers, all in one room working together in
• 43:10
  real time and that allows us to make decisions that are best for
• 43:14
  the whole car and to make them really fast.
• 43:17
  And that approach is unlike traditional Automotive
• 43:20
  Engineering, which is really fractured.
• 43:23
  And if you were to go by like a premium German electric car, the
• 43:26
  engineers who designed the drive inverter in that car.
• 43:29
  They did not work for that car company.
• 43:31
  They worked for a contractor and a Tesla, we design the entire
• 43:34
  car and the factory that builds it.
• 43:37
  I want to highlight a few examples of what we've been able
• 43:40
  to do in house. Thanks to that unique approach.
• 43:44
  So inside the charge of your Tesla, our transistor packages
• 43:49
  and that's at the top of your screen here.
• 43:51
  And every electron, that moves you down the road flows through
• 43:54
  one of these packages, we designed our own custom package
• 43:58
  which is what you're seeing here.
• 44:00
  And we can extract twice as much heat out of that package as what
• 44:03
  we could buy off the shelf. And so what does that mean?
• 44:06
  It means that the Silicon guard. The silicon carbide wafer that's
• 44:10
  inside those packages. Can be much smaller and silicon
• 44:13
  carbide. Died.
• 44:14
  It's an amazing semiconductor but it's also expensive, and
• 44:19
  it's really hard to scale. So using less of it is a big win
• 44:22
  for us. Then on top of that,
• 44:26
  orchestrating all of these transistors and making them
• 44:28
  switch in the right ways is computationally extremely
• 44:31
  intensive. It used to require for
• 44:34
  microprocessors, which are shown here.
• 44:36
  In the bottom left, we have developed our own custom
• 44:40
  microprocessor, its purpose built for high power
• 44:43
  Electronics. It's half the cost and it does
• 44:47
  in just one the job of all those four.
• 44:50
  And these are just two examples of many that I could use to
• 44:53
  showcase our expertise, in high power electronics.
• 44:56
  And that expertise is allowed us to take the cost of the
• 44:59
  Chargers, they were in our model.
• 45:01
  S, when we launched it in 2012, both the cost and the mass and
• 45:05
  cut both of those in half. And even more important, Power
• 45:10
  Electronics, their Central, not just to our cars, there are also
• 45:14
  Central to our superchargers into our energy storage
• 45:17
  products, and Rebecca, and Mike will be talking more about that.
• 45:21
  So, in addition to the work that we've done is software and
• 45:24
  Hardware. We've also done a lot of work in
• 45:26
  house on software. So this is the Drive Unit from 0
• 45:28
  to 3 and Y. And if we take a cross section,
• 45:31
  we see the stator and the rotor they're responsible for the core
• 45:34
  function of the Drive Unit, which is converting electricity
• 45:37
  into motion. And our custom software, lets us
• 45:41
  simulate the rotating. Magnetic field that is
• 45:44
  responsible for that conversion and getting that simulation
• 45:47
  exactly, right. It's Central to the cost, the
• 45:51
  weight, the size, and even the sound of the Drive Unit, Now you
• 45:56
  can buy software that will do all of this but our tools are
• 46:00
  faster and they are more accurate and that was not easy
• 46:03
  to do and that allows us to quickly iterate through millions
• 46:08
  of possible Drive Unit, designs to find the best one.
• 46:12
  Want to highlight one more area where Tesla really excels,
• 46:16
  because we integrate work. That is often farmed out.
• 46:20
  So when you are making a new product, it's not enough to
• 46:23
  think about the product itself, you have to think about how
• 46:26
  you're going to make it at scale.
• 46:28
  So Tesla are powertrain and or powertrain manufacturing.
• 46:31
  Equipment is both designed Under One Roof.
• 46:34
  The engineers were designing the motor.
• 46:36
  They are in the same room as the engineers who are designing.
• 46:38
  The machine that's going to put that motor together, and that
• 46:42
  go. Vibration pushes us from day.
• 46:44
  One to design products that are not only high performance, but
• 46:48
  are really easy to assemble. So all of its expertise that we
• 46:54
  have in the powertrain team and hardware and software in
• 46:57
  manufacturing, it's going to have a major impact on our next
• 47:00
  platform. In our next powertrain.
• 47:05
  So the silicon carbide transistors that I mentioned
• 47:07
  that are key component. But expensive, we figured out a
• 47:11
  way to use 75% less without compromising, the performance,
• 47:15
  or the efficiency of the car. And of course, we know the
• 47:19
  battery cells supplies, one of the constraints on the
• 47:21
  scalability of he's right now our new powertrain is compatible
• 47:25
  with any battery chemistry that will give us great flexibility
• 47:30
  in battery sourcing. If we want to make Eevee's more
• 47:34
  accessible to people that have to be cheaper, we've reduced the
• 47:38
  drive unit cost about $1000 and we don't think any other
• 47:42
  automaker is even close to that number.
• 47:47
  Finally, the bigger factor is the longer.
• 47:49
  It takes to build. If we can build the same number
• 47:52
  of cars from a smaller Factory, we are going to be able to
• 47:54
  scale. EV production faster.
• 47:56
  Our next powertrain Factory is 50% smaller than the one that's
• 48:00
  behind me today, even though it has the same capacity.
• 48:06
  All of these improvements, I think are going to be
• 48:08
  transformative for the adoption of Ibiza and our ability to
• 48:11
  scale them. There's one more thing that I
• 48:12
  want to highlight. So I talked about how we had
• 48:16
  reduced the amount of rare earth in our powertrains and as the
• 48:20
  world transitions to clean energy, demand for Rare Earth is
• 48:24
  really increasing dramatically. And not only is it going to be a
• 48:27
  little hard to meet that demand. But mining that rare earth it
• 48:31
  has environmental and health risks.
• 48:33
  So we want to do even better than this.
• 48:37
  We have designed our next Drive Unit.
• 48:39
  Chooses a permanent magnet motor to not use any rare earth
• 48:43
  materials at all. So how does all this fit into
• 48:48
  the master plan? We can make lower cost products
• 48:51
  that are still efficient and compelling and we can make them
• 48:53
  at scale. We're going to use less of
• 48:56
  constrained, Commodities, silicon carbide rarest.
• 48:59
  We're going to build them all in Compact and high output
• 49:01
  factories that are easy easy for us to build quickly.
• 49:06
  We're going to make that easy to scale powertrains all the way up
• 49:09
  to the levels of the Drew, anyone mentioned at the
• 49:10
  beginning. And this achievement like all of
• 49:13
  the achievements that I mentioned today, it's only
• 49:15
  possible because of the incredible people on our
• 49:18
  powertrain teams. They are absolutely committed to
• 49:21
  the cause of sustainable energy. And that is why we can do what
• 49:24
  no other company can do. Thank you.
• 49:35
  So next, I'd like to welcome my colleagues Pete.
• 49:37
  Bannon David Lowe. Thanks for calling.
• 49:45
  Good afternoon everybody and welcome to Texas.
• 49:48
  Pictured here is the low voltage system of a Model.
• 49:51
  S from 2012. This car was primarily designed
• 49:54
  with vendor Source controllers integrated by Tesla there are
• 49:58
  over 300 volt. Low voltage devices that are
• 50:01
  Compass everything from the mundane like the light in the
• 50:04
  glove box to the complicated like the infotainment computer
• 50:08
  to the safety-critical like the airbags steer-by-wire and break
• 50:12
  by wire. The low voltage harness is built
• 50:16
  from Individual wires cut to length crimped and inserted into
• 50:20
  connectors. A manual process that is tedious
• 50:26
  error-prone and doesn't scale well going forward.
• 50:30
  We want to reduce the size and complexity of the harness and
• 50:33
  enable automated manufacturing. These wire harnesses introduced
• 50:37
  extraordinary complications especially in the early stages
• 50:40
  of development. Bring up a vehicle because when
• 50:42
  we're trying to bring up this entire system, We see that it's
• 50:45
  not working properly. We don't know whether it's a
• 50:47
  problem with the software with the controller, the processor,
• 50:50
  any of the endpoints in this entanglement of wires and so we
• 50:53
  have to go and debug everything all at once and Pete's going to
• 50:56
  tell us about how we're going to make it better.
• 50:59
  So I'm going to talk briefly about a few things that we've
• 51:02
  changed over the years since the model S with the model 3 Tegra
• 51:06
  Tesla started designing an increasing share of our
• 51:08
  controllers by merging controllers together and we were
• 51:13
  able to simplifies the design. Sign with a significant
• 51:16
  reduction in wire count and weight going from.
• 51:19
  S 2 3. We reduce the wire harness by 17
• 51:22
  kg compared to Model S and this is a pretty big deal and just to
• 51:26
  put that into context, our VP of engineering Lars will deliver a
• 51:30
  bottle of your favorite Spirit to your desk.
• 51:33
  If you manage to save a kilogram of weight for my car, so this
• 51:36
  Improvement costume quite a lot. We Skip One.
• 51:42
  Yep. The original model 3 controllers
• 51:44
  were enhanced and then built into model.
• 51:47
  Why those improve controllers where then move back into model
• 51:50
  3 when we introduce the heat pump at that point we had a
• 51:54
  shared controllers across both cars which helps simplify our
• 51:57
  supply chain. Those model 3 and Y controller.
• 51:59
  Designs were then updated and enhanced for the new versions of
• 52:02
  Model. S, and X that we introduced in
• 52:04
  2021 for cyber truck, we are now.
• 52:07
  Designing 85% of the Hours in the car for the next-gen
• 52:11
  platform, we're going to finish the job and be designing. 100%
• 52:14
  of the controllers to give us full control of the design and
• 52:17
  the supply chain at the component level having full
• 52:21
  control of the supply chain. At the component level has been
• 52:23
  critical to Tesla over the last few years, as a supply
• 52:26
  constraints have hampered, our ability to build cars and with
• 52:31
  that control comes control over all the software, which enables
• 52:34
  us to develop features and functionality that we never even
• 52:38
  dreamed of at the time. We Design the hardware.
• 52:40
  So that's why you see software in your car is getting better
• 52:43
  and better over time in ways that we didn't even think of
• 52:45
  when we design the hardware in the first place.
• 52:48
  Yeah. Sentry mode is one of my
• 52:49
  favorite last-minute. Change my favorites with the
• 52:53
  intron Chuck introduction of model. 3 in 2017, we deleted
• 52:58
  relays and fuses from the car in favor of e fuses.
• 53:01
  He fuses replace moving Parts with solid state transistors,
• 53:04
  they provide fine-grained control of the power system to
• 53:08
  software and allow software. To do Advanced things like load
• 53:11
  shedding in the car, under adverse conditions.
• 53:14
  If uses also allow for software-controlled retries for
• 53:17
  transient faults and they all allow for detailed monitoring of
• 53:20
  the power system over time. And what the theme you hear a
• 53:24
  lot about today, is software, control, the hardware?
• 53:26
  And when I think about that in my team, software-controlled
• 53:29
  Hardware is fundamentally about being able to sidestep what
• 53:32
  otherwise would have been static trade-offs between one attribute
• 53:36
  or another in a piece of Hardware we are able with We're
• 53:40
  to instill, intelligence context, awareness, and context,
• 53:44
  specific behavior, into what otherwise would have been a
• 53:47
  piece of Hardware that had to get optimized for one type of
• 53:49
  scenario. We get more of everything in
• 53:53
  2022. We completed the transition of
• 53:55
  the lead-acid batteries to lithium-ion batteries and and
• 53:58
  which use a new tool list connector.
• 54:01
  We expect Lithium-ion batteries to last for the life of the car.
• 54:04
  So, don't expect anyone to wake up to a dead battery ever again,
• 54:09
  this will Emanates a major source of failures for our car
• 54:13
  and the tool is connector makes it easier to service the car and
• 54:17
  it also includes a software feature to allow validation that
• 54:22
  the connector has been properly installed at the end of a
• 54:24
  service event which removes another source of failures in
• 54:27
  our cars. Obviously the mass and volume
• 54:30
  savings are also significant which is super helpful.
• 54:37
  Tesla were always trying to improve every single component
• 54:40
  in the car and a nice example of that is the 15 inch display.
• 54:42
  That was originally shipped in the 2017 model 3 over time, the
• 54:48
  cost of the display has gone down 24%.
• 54:51
  We've been able to reduce the weight by 12% and we've reduced
• 54:55
  the power by 33 percent. At the same time, we've
• 54:58
  increased the brightness of the display by 50% and improve the
• 55:01
  color accuracy. So this is one of our favorite
• 55:03
  things at Tesla does make a component cheaper.
• 55:05
  Iran better. At Tesla, the drive to improve
• 55:09
  the efficiency of the car is never ending.
• 55:11
  I think of it as like peeling, a carrot.
• 55:12
  You just take a swipe at it, a little bit falls off.
• 55:15
  Doesn't really make much of a difference, but if you just keep
• 55:18
  at it over time, it accumulates into a pretty significant
• 55:21
  Improvement. One of the changes that we're
• 55:23
  looking to make right now is to change something that's been
• 55:26
  steady for the last 60 years. 12 volts. 12 volts has been, as I
• 55:33
  said for 60 years, the demand for power in the car has been
• 55:35
  steadily increasing to the point where Now, have to have pretty
• 55:38
  large wires to drive over 200 amps of current around the car,
• 55:42
  which increases the Mast and cost with cyber truck and all
• 55:45
  future. Tesla platforms, we will be
• 55:46
  moving to 48 volts. This reduces the current needed
• 55:50
  by a factor of four and since power loss in the harness is
• 55:53
  resistance times. The square of the current, a 4X
• 55:57
  reduction and current leads to a 16 x reduction in lost power,
• 56:01
  while Distributing energy in the car that allows for smaller
• 56:05
  wires smaller, e, fuses and smaller controller.
• 56:07
  Hours. It also allows us to make this
• 56:09
  heat sinks smaller or in many cases, remove it completely.
• 56:13
  All benefiting the car in terms of mass and weight 48.
• 56:18
  Volts is the future for a little voltage design at Tesla and
• 56:21
  likely the rest of the industry. In due course, we welcome and
• 56:24
  encourage other oems and the entire supplier Network to join
• 56:28
  us on this Evolution. You bet.
• 56:33
  The number of wires in the car is driven by the number of end
• 56:36
  points that need to be powered and controlled in the past.
• 56:39
  Centralized control has led four wires spanning the entire car
• 56:43
  for cyber Truck design. We have moved to a local
• 56:46
  controller where the wire is connected to the nearest
• 56:50
  controller and those controllers are connected over.
• 56:53
  Ethernet wires are routed to the nearest controller where the
• 56:56
  data is converted into a network packet for a transition
• 56:58
  transmission to the correct location on the car to be
• 57:02
  effective. The Work must be, reliable have
• 57:04
  low, latency and low Jitter. And these are all attributes
• 57:06
  that we've been able to achieve with the current design.
• 57:10
  The design is delimited, most of the Cross work, our wires and
• 57:13
  cyber truck. And with the next-gen platform,
• 57:15
  we're going to finish the job and eliminate all of them.
• 57:18
  The some Consolidated vehicle Network allows us also to make a
• 57:21
  whole lot more Dynamic changes on the fly to how components in
• 57:25
  the vehicle talk to each other rather than the traditional
• 57:28
  approach of separate can buses that are spread throughout the
• 57:30
  vehicle. Yeah, and West, it Hardware. and
• 57:33
  one of the nice things is that you can see the entire vehicle
• 57:35
  through a single connection which wasn't possible in the
• 57:37
  past, for debug, For the next-gen platform looking to
• 57:42
  optimize the controller design across the entire car and across
• 57:46
  all of the organizations not just for a given subsystem
• 57:49
  simplifying. The wire harness will enable
• 57:51
  automation 48 volts will allow us to reduce the size mass and
• 57:56
  cost of the low voltage system and is one of the key things to
• 58:00
  enabling us to scale production of the low voltage systems to
• 58:04
  for the master plan 3. David.
• 58:09
  Cool. So, I joined Tesla in 2012 and
• 58:13
  even in those very early days of the model, S, we already have
• 58:17
  the ability to update software over the air in all the
• 58:21
  controllers throughout the entire vehicle at the time.
• 58:23
  When other cars were just starting to update software in
• 58:27
  only their infotainment units. And furthermore, we've had
• 58:30
  sophisticated anonymized data logging and Telemetry
• 58:33
  capabilities that allow us to understand how the fleet is
• 58:37
  doing what customers are doing. With it and how it's responding
• 58:41
  in the field to that usage. These two capabilities combined
• 58:46
  OT a updates and Data Insights. Give us the ability to iterate
• 58:50
  quickly on our software and to maximize the amount that we
• 58:53
  learn and and proceed. And the progress that we make in
• 58:57
  every one of those iterations, we've used these capabilities to
• 59:01
  inform countless design decisions, in both software and
• 59:04
  Hardware. Yeah, one example is that we
• 59:07
  were able to Monitor and track the use of the sunroof and our
• 59:11
  cars and found that our customers never use their
• 59:13
  sunroof. So we've made it easy decision
• 59:15
  to remove it very barely. Yeah I've had some rest for a
• 59:18
  while. As another example, we are able
• 59:24
  to use our Collision data to design our crash safety systems
• 59:28
  for what happens in the real world far beyond what regulatory
• 59:32
  and consumer, ratings test prescribed and then further Our,
• 59:36
  we're able to use that data to recreate in simulation, all of
• 59:40
  those crashes, every single time we make a change to the vehicles
• 59:44
  design or to the software that controls the air bags and other
• 59:48
  restraint systems here on this graph.
• 59:52
  The gray dots that show up first are specific, crash tests that
• 59:55
  are prescribed by Regulatory and consumer ratings agencies
• 59:58
  throughout the whole world. And the rest of the color that
• 1:00:02
  fills in around and on top of it, those represent the crashes
• 1:00:06
  that we've seen happen in the real world.
• 1:00:08
  Those are what we design for. And what we test to write and
• 1:00:13
  last year, this data was used to change the algorithm for seat
• 1:00:16
  belt tensioning to reduce injuries in the field, which was
• 1:00:19
  0, ta to all of our cars. Here are a few statistics that
• 1:00:25
  demonstrate how staggeringly quickly we gain new insights
• 1:00:29
  from the anonymised data that we've received from the fleet
• 1:00:32
  everyday, everyday, 123 million miles driven 1.9 million charge
• 1:00:38
  sessions and those rates are only increasing as we deliver
• 1:00:42
  more and more cars into the world. as another example we've
• 1:00:48
  analyzed patterns in the way the fleet drives and charges to
• 1:00:51
  optimally sized battery packs for our next Generations of
• 1:00:54
  vehicles and you'll hear lots of other examples throughout the
• 1:00:57
  rest of the day about how we're using data to inform decisions
• 1:01:00
  in the product and in manufacturing, So you've heard a
• 1:01:04
  lot from Pete calling others about how important vertical
• 1:01:07
  integration is for us and Hardware, but it's especially
• 1:01:11
  important for us in software. And this is extremely difficult
• 1:01:15
  from the way. The rest of the traditional
• 1:01:17
  automotive supply chain is set up in most situations.
• 1:01:21
  All the controllers in the car are delivered by different
• 1:01:23
  tier-one suppliers whose software is written by tier two,
• 1:01:27
  suppliers who Farms some of that software out to different tier
• 1:01:31
  3, suppliers, and so on, and so on.
• 1:01:33
  I'm making a change that spans multiple components takes months
• 1:01:37
  of coordination before any work can even start.
• 1:01:41
  And furthermore, integration of software and the server side
• 1:01:45
  with what's happening in our vehicles.
• 1:01:47
  As always been a core part of our DNA and labels us to do
• 1:01:50
  things that nobody else can. Most Engineers think about the
• 1:01:54
  vehicle as a complete Consolidated, fully contained
• 1:01:57
  system. But in the software team, we
• 1:02:00
  think about the system as including the vehicle as well as
• 1:02:04
  all of our back-end server side, applications and infrastructure
• 1:02:07
  and the resultant feedback loop from the entire fleet that
• 1:02:10
  informs all the Ian's, we make as an engineering team.
• 1:02:15
  For example, we recently released a feature in model
• 1:02:18
  lesson X that automatically predictively adjusts and raises
• 1:02:22
  the suspension for ride Comfort before the car hits a section of
• 1:02:26
  rough road. We do that by leveraging.
• 1:02:29
  The fleet to generate a map of Road roughness everywhere, our
• 1:02:32
  cars drive, it sounds pretty simple, but requires
• 1:02:36
  coordination of software across a number of different components
• 1:02:38
  inside, and outside of the car, the restraint control module,
• 1:02:43
  whose inertial measurement. Unit senses the road roughness
• 1:02:46
  the autopilot computer whose GPS module and localizer figure out
• 1:02:50
  where the car is in the world, our navigation server which
• 1:02:54
  Aggregates the anonymize Telemetry from the fleet and
• 1:02:57
  annotates our map with things like updated speed, limits Lane
• 1:03:00
  topology. And now Road roughness, the
• 1:03:05
  onboard navigation engine, which looks ahead of the route that
• 1:03:08
  the car is traveling on and determines whether things are
• 1:03:11
  about to get rough. and finally, the air suspension controller,
• 1:03:15
  which takes all of those factors into account and decides
• 1:03:18
  continuously, whether it is appropriate to adjust the
• 1:03:22
  suspension for ride comfort, And before we release this feature
• 1:03:26
  to the world, we sent prototype versions of it.
• 1:03:29
  To the entire fleet that ran passively in the background.
• 1:03:33
  Sending us Anonymous data about every time it would have
• 1:03:37
  engaged, which taught us. Exactly how it was going to
• 1:03:40
  behave in the wild, in every environment in every
• 1:03:43
  circumstance of being able to test new algorithms in the
• 1:03:47
  background without impacting the car, as really critical ability
• 1:03:51
  for us, especially for safety critical, things like Samurai,
• 1:03:53
  automatic emergency. See breaking and we've used it
• 1:03:56
  to iterate on early versions of our stability control
• 1:04:00
  algorithms, which were introduced to the entire model 3
• 1:04:03
  Fleet in 2017 with extreme extremely high confidence, even
• 1:04:07
  on a very foundational, and safety critical function.
• 1:04:13
  Okay, so far you've heard me talk a lot about how we leverage
• 1:04:16
  our software systems to iterate quickly on the customer-facing
• 1:04:20
  product. But what I haven't talked about
• 1:04:22
  and what we haven't said much about publicly in general is how
• 1:04:25
  we leverage these capabilities in our own internal operations.
• 1:04:28
  So for example in the early days of manufacturing Model, S we
• 1:04:33
  very quickly realized that there are a lot of different ways you
• 1:04:36
  could miss assemble a car. You could forget to plug in a
• 1:04:39
  wire. You could not fully set a Actor,
• 1:04:43
  you could pull apart from the wrong PIN.
• 1:04:45
  So when we designed our manufacturing processes from
• 1:04:48
  model 3 and all of our other vehicles going forward, we took
• 1:04:51
  a page from the software Playbook which says, test early
• 1:04:56
  and test often, And we applied that Mantra to every single car
• 1:05:01
  that we build at every step of the assembly line.
• 1:05:05
  So now when a production associate plug something into
• 1:05:08
  the car, the central car computer sees that connection
• 1:05:11
  confirms that it's the correct part for the type of car that
• 1:05:14
  supposed to be built installs. A software update if necessary
• 1:05:17
  checks, configurations, and calibrations and runs a barrage
• 1:05:20
  of tests on that thing and all the other things that are
• 1:05:22
  connected to it. And if it finds an anomaly that
• 1:05:25
  can't be fixed purely with software it throws an alert
• 1:05:28
  immediately that is displayed prominently on the vehicles
• 1:05:31
  display as well as sent to back and command and control systems.
• 1:05:34
  So that a Human can come over and fix the problem before.
• 1:05:38
  The rest of the car gets built on top of it.
• 1:05:41
  And we're taking a similar approach in service through a
• 1:05:44
  combination of onboard vehicle diagnostics, natural language
• 1:05:48
  processing applied to customer narratives.
• 1:05:50
  When a customer schedules, a service appointment and a suite
• 1:05:53
  of internally developed tools. We are successfully, diagnosing
• 1:05:57
  scheduling and ordering parts for over 33 percent of customer
• 1:06:01
  concerns in service. Yep.
• 1:06:04
  And we also use testers built around our own controllers to
• 1:06:07
  test subassemblies before they get to the car, which will be
• 1:06:10
  even more important as we go to lars's unboxed assembly strategy
• 1:06:14
  under percent will know with extremely high confidence that
• 1:06:17
  before those boxes get, put together that they are that they
• 1:06:20
  are assembled. They themselves are assembled
• 1:06:22
  correctly. Okay, you've heard it lots of
• 1:06:27
  other events about all the awesome things that are
• 1:06:29
  autopilot and AI teams are doing to make the car drive itself,
• 1:06:33
  bringing us into the future of autonomy.
• 1:06:36
  But we've also been thinking for years about all of the other
• 1:06:39
  pieces that we are going to need to manage a network of
• 1:06:42
  autonomous vehicles. A lot of this has been happening
• 1:06:45
  behind the scenes in the form of platform level functionality,
• 1:06:49
  that will leverage later but you've seen some of its surface
• 1:06:52
  already in terms of features that our customers and our
• 1:06:55
  internal. Operations teams can benefit
• 1:06:56
  from like in 2021. We built on top of our mobile
• 1:07:00
  app, phone key and gave our customers, the ability to share
• 1:07:02
  their car with anyone, by sending them an invitation in a
• 1:07:05
  text message or an email last year, we introduced profile
• 1:07:09
  synchronization, which synchronizes your seat steering
• 1:07:12
  and mirror positions as well as your settings.
• 1:07:14
  Media, favorites and stored, navigation locations across all
• 1:07:18
  vehicles in your account. Internally, we have an app that
• 1:07:22
  allows our Personnel in engineering manufacturing,
• 1:07:25
  delivery and Logistics to view, locate and drive all Tesla on
• 1:07:30
  vehicles at their site. And when customers bring their
• 1:07:35
  car into service, when they receive a Tesla own service
• 1:07:38
  loner vehicle, we are starting at some locations to
• 1:07:41
  automatically add that vehicle to the customers mobile app
• 1:07:44
  account. So that combined with Cloud
• 1:07:46
  synchronized profiles and phone key, it's a completely seamless.
• 1:07:50
  Experience as soon as the customer walks up to that
• 1:07:52
  loaner, it behaves and feels exactly like their own car.
• 1:07:57
  And all this is built on top of end-to-end, encryption and
• 1:08:01
  cryptographically signed commands.
• 1:08:03
  So that customer data remains private and obscure and the
• 1:08:07
  fleet only trust commands from authorized parties.
• 1:08:11
  So that's a bit of a preview of all the pieces that were
• 1:08:14
  building looking ahead to our future, of autonomous fleets
• 1:08:17
  synchronization, permissions, management, security, and
• 1:08:21
  privacy. And that has been a whirlwind
• 1:08:26
  tour through some of the things that we're doing behind the
• 1:08:28
  scenes to enable the efficiency cost reduction and speed that we
• 1:08:33
  are going to need for their next phase of growth for the next
• 1:08:36
  phase of the master plan. And so now to tell us about the
• 1:08:39
  latest in full self driving, I'd like to introduce a joke.
• 1:08:52
  Thanks, David and Pete. Hey everyone, my name is,
• 1:08:54
  Jacques kallis for me. I work on autopilot and sell
• 1:08:57
  driving a Tesla. I joined Tesla back in 2014.
• 1:09:00
  So I've been working on this problem for almost nine years
• 1:09:02
  now. Yeah, some of you might be
• 1:09:06
  wondering, hey what the self-driving got to do with the
• 1:09:09
  plan to a sustainable future. But it's actually a critical
• 1:09:12
  part of this plan. And here's why Currently, when
• 1:09:17
  the car is not being used, it is sitting idly in.
• 1:09:20
  Parking lots. Not doing anything.
• 1:09:23
  But when autonomy is truly unlocked this car, instead of
• 1:09:26
  being idle can go serve other customers.
• 1:09:31
  This fundamentally. Now reduces the need to scale
• 1:09:36
  manufacturing to extreme levels because each car is being used
• 1:09:40
  way more. But this is no trivial problem.
• 1:09:46
  Building a scalable self-driving system is I think one of the
• 1:09:49
  hardest, real world are problems out there right now.
• 1:09:52
  Nonetheless, we at Tesla have made significant strides in
• 1:09:57
  making one of the most General systems at solving this problem.
• 1:10:06
  There are three main parts to get right to build a scalable
• 1:10:09
  problem. First is the architecture, the
• 1:10:13
  architecture of the air system, second is the data and third is
• 1:10:18
  the computer will start with the architecture.
• 1:10:24
  A Tesla, we are betting on AI machine, learning neural
• 1:10:27
  networks to help us build a general vision and planning
• 1:10:30
  system. In the early days we used to
• 1:10:33
  have single camera. Single frame neural networks,
• 1:10:36
  that produce some outputs. These Fest is together in some
• 1:10:40
  post-processing steps for the planner but this was very
• 1:10:43
  brittle and was not leading to great success.
• 1:10:46
  So what we did in the last few years is transition, most of our
• 1:10:49
  stack into this multi camera video.
• 1:10:52
  Unit works these neural networks take in the live feed from the
• 1:10:58
  car in real time of the eight cameras in the car.
• 1:11:02
  And produce a single unified 3D output space.
• 1:11:08
  There are many tasks that we produce such as the presence of
• 1:11:11
  obstacles that motion Lanes roads traffic lights.
• 1:11:15
  What have you? This is one example output that
• 1:11:19
  you're seeing here. This is from our occupancy
• 1:11:21
  Network that predicts it positions of obstacles and then
• 1:11:24
  motion you can see that it precisely captures.
• 1:11:27
  These were we violent motion of this truck next to us and this
• 1:11:31
  helps the planning system to avoid a collision with this
• 1:11:33
  object. Some of the tasks such as Lane
• 1:11:39
  connectivity or more complicated to model using methods in
• 1:11:43
  computer vision. So we don't stop at the computer
• 1:11:46
  vision techniques but we reach out to techniques in other
• 1:11:49
  areas, such as language modeling reinforcement, learning to model
• 1:11:53
  this task. This is similar to these
• 1:11:56
  networks use similar techniques such as Transformers.
• 1:11:58
  Attention modules, autoregressive modeling of
• 1:12:02
  tokens, similar to what large long is models such as GPD do of
• 1:12:07
  there. With such an end-to-end system
• 1:12:10
  of solving perception. We have really removed the
• 1:12:14
  brittle post-processing steps and produce, high quality
• 1:12:16
  output, for the planning system. Even the planning system is not
• 1:12:22
  stuck. In the old ways, it is now
• 1:12:24
  starting to use more and more AI systems to solve this problem.
• 1:12:28
  The neural network based planners are needed.
• 1:12:31
  Especially in complicated, urban planning when there's a lot of
• 1:12:34
  other objects interacting with us.
• 1:12:36
  This is, for example, an intersection where we have to
• 1:12:38
  turn left while eating two Crossing objects and to the
• 1:12:41
  pedestrians crossing the road, they have to do this, both
• 1:12:44
  safely and smoothly while respecting everyone's
• 1:12:47
  right-of-way and preferences. If this was done, namely, each
• 1:12:53
  configuration would take 10 milliseconds of compute and are
• 1:12:56
  easily easily thousands of configurations to reason about
• 1:13:01
  this would not be feasible using traditional compute but using AI
• 1:13:04
  we have packaged all of this into a 50 millisecond compute
• 1:13:08
  budget, so it can run real time. The second big piece of this
• 1:13:14
  puzzle is the data, this is where test as a unique Advantage
• 1:13:18
  because it can tap into the feet, tap into the fleet to
• 1:13:22
  access the exact data that can fix the problems.
• 1:13:28
  But raw data is not sufficient. Not a train.
• 1:13:32
  This networks, you need label data.
• 1:13:35
  You need labels to supervise the networks.
• 1:13:39
  And if you only depended on the human laborers, this data would
• 1:13:43
  be too tiny to train. These large multi camera video
• 1:13:47
  modules. When you lat, lots and lots of
• 1:13:49
  data to train these networks. Hence, we have built a
• 1:13:53
  sophisticated auto-leveling pipeline That collects data from
• 1:13:57
  the fleet runs computation algorithms in our data center
• 1:14:02
  and then produces the labels to train these Networks.
• 1:14:06
  Here you are seeing a 3D reconstruction that is
• 1:14:09
  happening. By collecting various clips from
• 1:14:11
  different cars in the fleet and assembling them all of them
• 1:14:14
  together into a single unified representation of the world
• 1:14:17
  around the car. You can see all the lanes, the
• 1:14:20
  road boundaries curb crosswalks in the text on the road being
• 1:14:23
  accurately. Reconstructed by this algorithms
• 1:14:30
  You don't have to stop at this kind of reconstruction.
• 1:14:33
  Once we have this based reconstruction, we can build
• 1:14:36
  various simulations on top of it to produce an infinite variety
• 1:14:40
  of data to train. For all the corner cases, we
• 1:14:44
  have a capable simulator that can synthesize adversarial,
• 1:14:48
  whether lighting conditions and even the motion of other objects
• 1:14:52
  to test all the corner cases. That might even be rare in the
• 1:14:55
  real world. Here's an example of why this
• 1:15:00
  data is critical and how we can solve problems using data.
• 1:15:04
  So back in the days for example, in this case, we had some false
• 1:15:09
  breaking where we thought this car.
• 1:15:11
  That's actually part there was going to move into our path and
• 1:15:14
  hence, we are pre cautiously breaking But it is unnecessary
• 1:15:19
  because the car is truly part. So how can we solve this?
• 1:15:23
  What we did was we mined the fleet for similar cases, where
• 1:15:27
  the current Falls breaking due to some parkour.
• 1:15:31
  The added 14,000 videos to our training set.
• 1:15:33
  And then once we train the Network's again with this new
• 1:15:36
  data, it now correctly, understands that okay?
• 1:15:39
  That is no driver in this car and it must be part.
• 1:15:41
  So then there is no need to break and this is a systematic
• 1:15:45
  way to solve problems on the chart.
• 1:15:47
  On the right side, you can see that every time we add data,
• 1:15:50
  they performance improves, and then we can do this for every
• 1:15:54
  kind of task that we have in our system.
• 1:15:57
  And this is what we call as data engine.
• 1:16:00
  We identify challenging cases such as the one you saw earlier,
• 1:16:04
  there could be other different types of challenging cases to we
• 1:16:07
  mine the fleet for such data, Put it through an auto leveling
• 1:16:11
  system and produce the labels added to the training set.
• 1:16:15
  And once you have the newly trained models, we departed unto
• 1:16:18
  the Feet Fleet, If we rinse and repeat this process, everything
• 1:16:24
  gets better and better. The final critical piece is the
• 1:16:31
  compute. In order to train, these large
• 1:16:33
  models in a reasonable amount of time.
• 1:16:35
  You need lots of compute. In addition, computer is also
• 1:16:38
  required to produce the labels automatically.
• 1:16:41
  This is just compute in our data centers.
• 1:16:43
  In addition, we also have high compute computers in the car
• 1:16:50
  which can run up to. I want 50 drops of compute.
• 1:16:55
  On the back end, we have 14,000 GPU cluster and roughly. 30
• 1:16:59
  percent is used for auto labeling and the remaining 70
• 1:17:01
  percent is used for training. We also have a 30 petabyte of
• 1:17:07
  video cash and this is growing to 200 petabytes.
• 1:17:10
  All of this is going to significantly increase one.
• 1:17:13
  We once we bring Dojo, which is our training computer onboard
• 1:17:17
  into this, just to given reference, just the occupancy
• 1:17:20
  networks that you saw earlier. Use one point, four billion
• 1:17:23
  frames to train these Networks. We have already shipped out
• 1:17:31
  fsdb, the software to pretty much everyone that has bought it
• 1:17:34
  in United States and Canada this amounts to roughly 400,000
• 1:17:38
  customers who can turn it on anywhere and then the car would
• 1:17:41
  attempt to drive to the destination is to supervise but
• 1:17:45
  it can already handle turns stop at traffic lights yield with
• 1:17:49
  other objects and generally get to the destination.
• 1:17:53
  We are observed that people who use FSD or already 526 safer
• 1:17:58
  than the user u.s. national average.
• 1:18:01
  Like I mentioned earlier the solution to scalable FST is
• 1:18:05
  getting the a architecture, the data and the compute just right.
• 1:18:10
  And we are assembled a world-class team to execute on
• 1:18:13
  this. You're pushing the frontier on
• 1:18:15
  these on these three items and as we improve the safety, the
• 1:18:19
  reliability and the comfort of a system.
• 1:18:21
  We can then unlock Rivals operations, which then makes the
• 1:18:25
  car be use way more than what it is used right now.
• 1:18:30
  That I'd like to introduce you on back on stage to give some
• 1:18:34
  more updates. Next video.
• 1:18:54
  It's going to wear today in the arms and legs just separate.
• 1:19:04
  We have a whole lap full of arms and legs.
• 1:19:17
  With bearing in mind, that when we did AI de this first box, you
• 1:19:23
  must didn't work. Walk at all.
• 1:19:26
  So the rate of improvement here I think is quite significant.
• 1:19:30
  It's obviously not doing parkour, but it is walking
• 1:19:36
  around and we have multiple multiple copies, I suppose of
• 1:19:41
  Optimus the thing that's I think Tesla brings to the table that
• 1:19:52
  others don't have is that we have We have the real world AI.
• 1:20:03
  We're the most advanced in real-world AI so the same AI
• 1:20:07
  that drives the car which you can think of the car.
• 1:20:14
  Really is a robot on Wheels. And this is a robot on legs.
• 1:20:18
  So as we solve real-world a I don't think there's any I don't
• 1:20:22
  think there's anyone even close to Tesla on solving real-world
• 1:20:25
  AI that same computer and software.
• 1:20:30
  R goes into optimist. So it's not that helpful to have
• 1:20:35
  a humanoid robot. If you have to program every
• 1:20:38
  individual action, it needs to be able to walk around
• 1:20:42
  autonomously and solve tasks. You should be able to instruct
• 1:20:46
  it in simple. Things by showing visually what
• 1:20:50
  your, what the robot needs to do or just telling it what to do
• 1:20:54
  so. So, I think that's a key
• 1:20:57
  advantage that we have. And then, we also, Good at
• 1:21:01
  designing things for manufacturing and then
• 1:21:03
  manufacturing itself. So the the actuators in Optimus
• 1:21:08
  are all custom design, Tesla actuators.
• 1:21:11
  So we designed the like the electric motor.
• 1:21:13
  The gearbox, the Power Electronics, obviously, the
• 1:21:16
  battery pack, everything else that goes into Optimist.
• 1:21:21
  We're actually quite, we were quite surprised to find how
• 1:21:24
  little was available off the shelf because there's a lot of
• 1:21:29
  vast number of Electric motors gearboxes and whatnot that are
• 1:21:33
  available in the world and we found none of them were useful
• 1:21:37
  in a humanoid robot virtually none.
• 1:21:40
  So you have to Custom Design the actuators for a humanoid robot
• 1:21:51
  and so the same team that designed the groundbreaking
• 1:21:56
  electric motors that are in the say the model has plaid designed
• 1:21:59
  the the actuators in the robot. So I mean for practical
• 1:22:06
  purposes, what this means is that we should be able to bring
• 1:22:10
  an actual product to Market at scale that is useful.
• 1:22:15
  Far faster than anyone else. And you know, assuming that
• 1:22:20
  things I'm saying are true or at least put it.
• 1:22:24
  I think they are true. Can just, it's just a question
• 1:22:26
  of the timing. You start getting into
• 1:22:30
  interesting questions of like, what's the ratio of humans to
• 1:22:33
  humanoid robots? I think it might be greater than
• 1:22:36
  1 to 1. You know, because you could see,
• 1:22:40
  you could sort of see a use a home.
• 1:22:42
  Use for robots. Certainly industrial uses for
• 1:22:46
  robots, humanoid robots, I think I think we might exceed a
• 1:22:51
  one-to-one ratio of humanoid robots to humans.
• 1:22:59
  It's not even clear what an economy means at that point.
• 1:23:01
  You know, since it economy is output per person, times
• 1:23:05
  persons, but if output is much higher and there's no limit on
• 1:23:09
  person's then what's the actual limit on the economy, you know?
• 1:23:17
  We're still pretty far from kardashev scales here, but we're
• 1:23:22
  getting there. So, anyway, so probably the
• 1:23:28
  least understood or appreciated part of what we're doing at
• 1:23:30
  Tesla but will probably be worth significantly.
• 1:23:33
  More than the car side of things long-term.
• 1:23:39
  Charging. Hi there.
• 1:24:00
  My name is Rebecca to Nucci and I lead our charging
• 1:24:03
  infrastructure teams here at Tesla.
• 1:24:06
  And at Tesla charging, we have understood since day one, that a
• 1:24:09
  great charging experience is the linchpin to electric vehicle
• 1:24:13
  adoption. And that understanding has meant
• 1:24:15
  that we've always taken a holistic approach to charging.
• 1:24:18
  That's a word that you've heard a lot here today but what it
• 1:24:21
  means for charging is that we've considered every use case.
• 1:24:24
  We think a lot about what it means to charge at home, even if
• 1:24:27
  that home is an apartment or condo and we spend a lot of time
• 1:24:30
  thinking about, what does it mean to charge away from home?
• 1:24:33
  Including if that's for daily commuting, or if you're going on
• 1:24:35
  a road trip. If in this holistic approach has
• 1:24:38
  led to some pretty incredible results in 2022, we provided
• 1:24:42
  nine terawatt hours of charging across our various charging
• 1:24:46
  methods over. 50% of that was supplied via convenient, AC
• 1:24:50
  home, charging. And when our customers are away
• 1:24:53
  from home, they can visit one of our 80,000 charging points that
• 1:24:57
  includes 40,000 of our beloved superchargers.
• 1:25:02
  Now getting here has meant that we've spent ten years building
• 1:25:05
  charging infrastructure when basically no one else in the
• 1:25:07
  industry would do it. And while we certainly have a
• 1:25:09
  lot of areas that we want to improve those 10 years have
• 1:25:12
  afforded us the opportunity to get pretty good at charging.
• 1:25:17
  First, we have the industry's lowest deployment costs.
• 1:25:20
  Are costs are often 20% if not seventy percent lower than
• 1:25:24
  Alternatives. And that goes for both our
• 1:25:26
  supercharging hardware and deployments and our AC charging
• 1:25:30
  product lines. A lot of reasons that were able
• 1:25:32
  to achieve this, you've heard a lot of them here.
• 1:25:34
  Today, we're vertically integrated, we manufacture an
• 1:25:36
  engineer, all of our own charging equipment, we share
• 1:25:39
  components across our different product lines and on the
• 1:25:43
  supercharger side. We also install and operate all
• 1:25:47
  All of our own sites and that's led us to be pretty obsessed.
• 1:25:50
  With finding new ways to innovate around installation.
• 1:25:55
  As an example we have recently extended our excellence in
• 1:25:58
  manufacturing to how we build supercharger sites.
• 1:26:01
  We are pre building for post supercharger units at our
• 1:26:05
  Factory and giggle at our Factory in New York.
• 1:26:08
  We load them on a truck, we track him to site, and then we
• 1:26:10
  crane them into position. This method saves us.
• 1:26:14
  Fifteen percent on our deployment costs and we can
• 1:26:16
  install a site in a matter of days.
• 1:26:21
  Once we install a site, we also operate it really efficiently
• 1:26:25
  over the last few years. We've been able to cut our per
• 1:26:27
  kilowatt-hour cost by 40% again, a lot of reasons for this but
• 1:26:32
  one of the top of the list is we focused on increasing our site,
• 1:26:35
  utilization site, utilization is just how many sessions are
• 1:26:38
  kilowatt hours. Can we push through a site or a
• 1:26:41
  post and basically doing that allows us to spread our cost
• 1:26:45
  over more sessions? Thus lower cost per kilowatt
• 1:26:48
  hour, but of course that's easier said.
• 1:26:50
  Said than done, because when we push up side efficiency, of
• 1:26:53
  course, the risk is that we have a poor customer experience.
• 1:26:56
  We have wait time at our sites. This is where trip planner comes
• 1:27:01
  in trip planners are in vehicle routing or navigation system.
• 1:27:05
  Other electric vehicle manufacturers will, some of them
• 1:27:08
  have vehicle side data, other infrastructure providers have
• 1:27:12
  site data but it Tesla, we have both.
• 1:27:15
  And what that allows us to do is to use trip planner to Route
• 1:27:18
  Vehicles towards available sites and away from congested sites.
• 1:27:22
  So we can balance our utilization without risking,
• 1:27:24
  wait time and the results speak for themselves over the last few
• 1:27:28
  years, we've been able to Of up site utilization by thirty
• 1:27:31
  percent that means lower per kilowatt-hour cost, while also
• 1:27:34
  cutting our wait time in half and we think this can get even
• 1:27:38
  better today, we're feeding trip planner with real-time site
• 1:27:41
  information and information about Vehicles.
• 1:27:43
  Currently at sites going forward will be moving that to
• 1:27:47
  projecting site occupancy based on the understanding of what
• 1:27:50
  vehicles are currently routing to those sites.
• 1:27:53
  And ultimately our vision for trip planner is that it's the
• 1:27:56
  air traffic controller for electric vehicle charging a
• 1:27:59
  Cross all infrastructure on a global basis.
• 1:28:04
  And the last thing we focused on to get here is quicker, charge
• 1:28:06
  times, we're really proud of what we've been able to
• 1:28:09
  accomplish here. We've shaved off 30% of our
• 1:28:12
  charge time average supercharging site time, visit
• 1:28:15
  time over the last few years. This is taken improvements on
• 1:28:19
  the hardware side with our software, on our vehicles and on
• 1:28:22
  our infrastructure and we're really excited to continue
• 1:28:25
  pushing this trend down. Looking forward.
• 1:28:30
  The job does not get any easier. As you heard earlier, we aspire
• 1:28:35
  to a fully electrified Global Fleet that Fleet from an
• 1:28:39
  industry standpoint needs 9, PW hours of charging on an annual
• 1:28:44
  basis and while Tesla charging certainly doesn't have to supply
• 1:28:47
  all of that. It does require that for our
• 1:28:50
  part. We have a few new focuses in
• 1:28:52
  order to scale. First, if you want to fully
• 1:28:57
  electrified Global Fleet all of those vehicles have to have a
• 1:29:00
  great and reliable charging experience.
• 1:29:03
  So, as many of you know, recently we have started opening
• 1:29:06
  opening up our networks on a global basis over 50% of our
• 1:29:11
  superchargers in Europe are currently open to other electric
• 1:29:14
  vehicles. We've also opened up in Asia
• 1:29:16
  Pacific with our first sites in Australia and just yesterday we
• 1:29:21
  opened our first 10 supercharger sites here in in North America
• 1:29:25
  in the u.s. to other electric vehicles.
• 1:29:27
  We've also invested a lot to be able to make this a really easy
• 1:29:31
  experience for our new customers.
• 1:29:33
  All you have to do is sign into our Tesla app to unlock a post
• 1:29:36
  and start charging. We've also on physical site
• 1:29:40
  changes. We've also added Hardware to our
• 1:29:41
  sites, they're called Magic docks.
• 1:29:44
  And basically, that is installed in areas where we have different
• 1:29:46
  charging standards and they allow for other electric
• 1:29:49
  vehicles to come to our site without having to bring their
• 1:29:51
  own adapters or Hardware. We've also Just started
• 1:29:55
  installing our fourth generation, supercharger posts.
• 1:29:58
  Those are being installed in Europe first.
• 1:30:00
  And while it's not a big mention, not a big thing to
• 1:30:03
  mention. They do have longer charging
• 1:30:04
  cables so that we can more easily reach the charge ports of
• 1:30:07
  different vehicles. The second thing we need to make
• 1:30:11
  sure we do. When we talk about scaling all
• 1:30:13
  of this charging infrastructure is we need to make sure that
• 1:30:16
  it's powered from renewable sources.
• 1:30:19
  We're very proud that over the last two years, we procured
• 1:30:22
  enough renewable energy to offset the amount of charging
• 1:30:26
  we've provided to customers. But as we look forward and we
• 1:30:28
  talked about this fully electrified Fleet on a global
• 1:30:31
  basis. We really want to make sure that
• 1:30:33
  the demand for charging more closely follows when renewable
• 1:30:37
  sources are available. Now this chart is on an
• 1:30:40
  aggregate basis for the us and it varies based on what your
• 1:30:43
  renewable generation sources are locally.
• 1:30:45
  But basically what this shows is we want to kind of move that
• 1:30:49
  charging curve and for solar and wind in this example it means
• 1:30:53
  more daytime charging and we think the best way to go towards
• 1:30:57
  daytime charging is to consolidate into install AC
• 1:31:01
  charging. That is convenient and low-cost
• 1:31:04
  everywhere. Vehicles are typically parked
• 1:31:06
  throughout the day so that's really the plan.
• 1:31:08
  Well, Supercharging and home charging, definitely stay a big
• 1:31:11
  part of the puzzle. Our teams are currently scaling
• 1:31:14
  to install, AC charging everywhere Vehicles charge
• 1:31:17
  during the day. So we can power them from a
• 1:31:19
  renewable sources. So summing it up while we've
• 1:31:25
  spent ten years installing infrastructure.
• 1:31:27
  That is low cost, its efficient and it provides a great customer
• 1:31:30
  experience. We're really just getting
• 1:31:33
  started. When we talk about supporting
• 1:31:34
  master plan, part 3, we've got a scale capacity on an industry
• 1:31:38
  wide basis to 9. PW h on an annual basis.
• 1:31:43
  We've got to open up to non Tesla's and make the charging
• 1:31:45
  experience for everybody really great.
• 1:31:48
  And we've got a power, all of this via renewable sources.
• 1:31:53
  And one more thing. Yes, we've got a scale.
• 1:31:57
  We've got a scale, our infrastructure.
• 1:31:58
  And yes we want to power it via renewable sources but we're
• 1:32:01
  Tesla. So we also want to make sure
• 1:32:03
  that we're continuing to focus on providing really incredible
• 1:32:07
  charging experiences. With that want to welcome up our
• 1:32:11
  leaders of supply chain, Roshan and carne to talk a little bit
• 1:32:14
  about that topic. Thanks Rebecca.
• 1:32:30
  Hi. My name is Karen and my team
• 1:32:33
  leads supply chain for electronics, powertrain and
• 1:32:37
  Battery, a Tesla. And we've also got
• 1:32:39
  responsibilities for indirect purchasing construction,
• 1:32:42
  procurement, and warehousing and distribution.
• 1:32:45
  Hi. My name is Roshan.
• 1:32:46
  I've been with Tesla for over 14 years.
• 1:32:49
  My team manages vehicle Commodities solar Logistics
• 1:32:54
  planning and Capital Equipment. Ten years, forget that.
• 1:33:00
  So today we wanted to provide you a quick overview of Tesla
• 1:33:02
  supply chain and what really separates us from the typical
• 1:33:06
  automotive supply chain. As you've seen in presentations,
• 1:33:09
  prior from Pete and ruin some others, Tesla designs.
• 1:33:13
  A lot of the sub components that go into our vehicles.
• 1:33:15
  So we're not buying things that are off the shelf from
• 1:33:19
  suppliers. So Supply, chains not purely a
• 1:33:21
  commercial relationship as it is and most other companies we
• 1:33:24
  actually have an armed with in supply chain called Supply.
• 1:33:27
  And Civilization engineering and the burden of taking a drawing,
• 1:33:32
  a design from concept and turning it into thousands of
• 1:33:35
  products produced at the right cost.
• 1:33:37
  And at the right yield Falls with this group, of course, they
• 1:33:39
  work very closely with design engineering, but really the
• 1:33:43
  level of detail that our engineers get involved in terms
• 1:33:46
  of building the capability of the suppliers is pretty
• 1:33:49
  detailed. So, you know, once we get a new
• 1:33:52
  part naturally with electric vehicles, a lot of the supply
• 1:33:56
  chain doesn't exist Our Engineers are basically take the
• 1:33:59
  drawing, turn it into the manufacturing concept, do the
• 1:34:02
  equipment selection and physically, go to the supplier
• 1:34:05
  and stay there for weeks or months.
• 1:34:07
  However, long it takes to basically, do the line, bring
• 1:34:09
  up, once the line is brought up and we're hitting the the right
• 1:34:13
  rate, the right quality at the right yield, then we work on
• 1:34:16
  things like Automation and and Olympic movement and and those
• 1:34:19
  types of activities. So there's a brick, big group of
• 1:34:23
  mechanical. Electrical industrial you name
• 1:34:25
  it Engineers within supply chain.
• 1:34:27
  And it's their responsibility to take care of this.
• 1:34:30
  Now, this is a huge strategic Advantage because we manage
• 1:34:35
  every detail of our supply chain because we're so integrated with
• 1:34:38
  our partners that we know about issues before they happen, it's
• 1:34:42
  almost live. We've got dashboards to got
• 1:34:44
  connections and across the tent, you know, 10,000 or so factories
• 1:34:47
  across the world that are made managed making our components.
• 1:34:50
  We have a pretty good view of the health of those suppliers
• 1:34:53
  and this has been a key asset and how we've been able to kind
• 1:34:56
  of manage things. Through a lot of different
• 1:34:59
  issues over the past couple of years.
• 1:35:01
  So You know, supply chain is a game of perfection, Perfection
• 1:35:07
  is a passing grade. It's like a, you know, if one
• 1:35:10
  part doesn't show up that line goes down and our CEO finds out
• 1:35:14
  about it in less than 20 minutes.
• 1:35:15
  It's a Tesla is very different than Apple.
• 1:35:17
  They outsourced manufacturing. We do it in-house so Supply
• 1:35:20
  chains kind of right in the middle.
• 1:35:21
  And and Perfection is, is increasingly Elusive, and very
• 1:35:25
  expensive if you kind of look at all the the issues we've had to
• 1:35:31
  encounter with the past year and today.
• 1:35:33
  Just going to talk about how we got through the past.
• 1:35:37
  And then what are we going to do going forward?
• 1:35:39
  And how are we going to get there?
• 1:35:41
  So let me draw your attention to this.
• 1:35:43
  Slide on the left, you will see a portfolio, our existing
• 1:35:47
  portfolio of products, right? At the top, you've got the snx
• 1:35:50
  platform and the middle, you've got three.
• 1:35:53
  And why? I think everyone's more than
• 1:35:54
  familiar with those and then the Tesla energy platform, which is
• 1:35:59
  a very high rate of growth for us and usually, Forget the solar
• 1:36:04
  and supercharged and solar and supercharger.
• 1:36:06
  Thank you for Sean and you'll see a reduction in.
• 1:36:10
  So and then we basically got the Tier 1 and tier 2 parts.
• 1:36:14
  This might be boring. But Tier 1 supplier is any
• 1:36:18
  supplier that builds and supplies Parts directly for
• 1:36:21
  consumption in our five factories and our tier 2
• 1:36:24
  supplier, basically does the same thing for a tier 1
• 1:36:26
  supplier. So they are the supplier to our
• 1:36:28
  supplier. And depending on the supply
• 1:36:30
  chain of any one of these products.
• 1:36:32
  This can actually go Down to the tier 6 levels, for example, if
• 1:36:35
  you're talking about the battery.
• 1:36:36
  So we've really get involved with the minds of like what
• 1:36:40
  we're procuring, the lithium the Cobalt and all that sort of
• 1:36:43
  stuff, because we're preparing such high volumes of it.
• 1:36:45
  So you'll see a reduction in the number of Parts on the snx
• 1:36:50
  platform to the 3 and Y platform.
• 1:36:53
  And this is because of the great work, our design Engineers have
• 1:36:55
  done of making the car simpler and easier to manufacture in
• 1:36:59
  parallel. We actually had a similar
• 1:37:02
  strategy and supply chain. We had cast a wide net to a lot
• 1:37:05
  of suppliers for the s&x platform.
• 1:37:07
  There's a lot of suppliers of be invited to partner with us and
• 1:37:11
  we used that platform as sort of a filter to see which suppliers
• 1:37:16
  have the technical financial and cultural capabilities to match
• 1:37:21
  us. Right?
• 1:37:21
  Who do we invite to the party when we really scale things up
• 1:37:25
  and we, you know, we've really dramatically reduce the number
• 1:37:29
  of suppliers that we were dealing with for the 3 and Y
• 1:37:31
  platform and this is an approach will You need to do for future
• 1:37:34
  platforms as they come along, that are even more high volume.
• 1:37:37
  So this this is really been our approach.
• 1:37:40
  It's made things a lot easier and it, you know, we could fix.
• 1:37:45
  We could figure out the suppliers that were capable of
• 1:37:47
  moving, it or paste. And then the ones that didn't
• 1:37:50
  have the desire to, or the capability to you need, both
• 1:37:52
  desire and capability. Secondly, the complexity on the
• 1:37:57
  tier 2 level is very high, right?
• 1:38:00
  The, these are the suppliers to our suppliers and they There are
• 1:38:04
  a lot of components think about silicon, think about resistors
• 1:38:06
  capacitors, diodes all the little assemblies materials raw
• 1:38:11
  materials. Although that's even further
• 1:38:12
  down. Yeah, but all these little
• 1:38:14
  subassemblies that have to come in and Tesla.
• 1:38:16
  The, the complexity is a mess. So even though at the tier one
• 1:38:20
  level, we're talking about 8,000 Parts total.
• 1:38:23
  It seems like a lot but it's not a lot.
• 1:38:25
  The management of the tier 2 is really where we Excel.
• 1:38:28
  And I'd like to illustrate that with an example.
• 1:38:32
  Meet the car computer, very innocent sounding name and it's
• 1:38:36
  anything. But this is an absolute monster
• 1:38:39
  to manage. So I think you've all seen the
• 1:38:42
  picture on the left off the autopilot board.
• 1:38:44
  That's the top of the autopilot board.
• 1:38:46
  The autopilot board also has a bottom and the bottom is
• 1:38:49
  populated heavily with components on the other side of
• 1:38:52
  the heat sink. You've got the MCU, the
• 1:38:54
  multimedia cluster board, this board is equally complex, it's
• 1:38:58
  also double-sided 8 layered and these boards these computers
• 1:39:02
  Run. So hot at Peak operation, that
• 1:39:03
  they have to be liquid, liquid cooled through a heat sink.
• 1:39:06
  So this assembly requires taking those two boards and then
• 1:39:10
  bonding them to a heat sink, that's hermetically sealed.
• 1:39:14
  As an assembly of course, flashing the software and all
• 1:39:17
  that sort of stuff and then that's one part number that
• 1:39:20
  comes to Tesla. So this is an example of one
• 1:39:23
  Tier 1 part number. That's a very complex assembly
• 1:39:26
  to manage of the tier 2 level and there's more than 7,000
• 1:39:30
  components here. There's There's a, you know, as
• 1:39:33
  we stand here, a components being assembled onto a car
• 1:39:35
  computer, every one point four milliseconds, the lion that
• 1:39:39
  builds this computer is the length of a football field.
• 1:39:42
  So it's quite complicated. And initially when we first
• 1:39:45
  started building, this board, due to the complexity of it, we
• 1:39:47
  had to rely heavily on labor, but once we dial then the
• 1:39:50
  quality, the rates and the yield, we started focusing on
• 1:39:54
  making this more efficient, the only way to control cost is by
• 1:39:56
  removing labor. The first step is removing
• 1:39:58
  labor. The second step is fully
• 1:40:00
  automating. The third step is turning off
• 1:40:01
  the And letting the factory run ideally and that's going to be
• 1:40:04
  the goal for us here. So 95%, but our work is not
• 1:40:07
  done. We're going to be going a lot
• 1:40:09
  further than this. So This illustrates the point
• 1:40:12
  that a part is not a part, A Part has a lot of complexity
• 1:40:16
  underneath it supply chain as a game of multiple tiers and and
• 1:40:20
  what's made us successful is our involvement and all the details
• 1:40:24
  with our supplier Engineers that are Tesla employees their Tesla
• 1:40:27
  Supply Engineers. So they're like on our payroll
• 1:40:29
  that go on ramp up these capabilities that our suppliers
• 1:40:32
  And then just managing each and every attribute of it, you know,
• 1:40:35
  seven thousand a day through the chip, shortage through the
• 1:40:38
  pandemic, through all the other stuff that we had to deal with
• 1:40:41
  was was very difficult to manage but because we had all the
• 1:40:44
  details, we were able to pull this off.
• 1:40:46
  Of course, supply chain is not just about Parts, it's also
• 1:40:50
  about logistics and a lot of other things, so, handed off to
• 1:40:52
  Roshan for sharing some stats. Yeah, so talk about inboard
• 1:40:58
  complexity just over the coast course of just.
• 1:41:02
  Last year we moved about 16 million pallets from our
• 1:41:06
  suppliers to our factories, just to visualize that.
• 1:41:10
  If you put those pallets side by side it could cover the half the
• 1:41:15
  circumference of the Earth. And talking about again the 22
• 1:41:20
  complexity over 1 billion, electronic components get moved
• 1:41:25
  every week to support the production that content team and
• 1:41:29
  addresses. Right now 1 billion components
• 1:41:31
  week, right? 82 billion eight.
• 1:41:33
  This is going to go up by the way.
• 1:41:34
  And and on top of that, we also have the responsibility to make
• 1:41:38
  sure the right component is supplied to the right service
• 1:41:42
  location to support the growing Fleet of our vehicles.
• 1:41:46
  And we have about 600 And 85 service centers that these
• 1:41:50
  components need to show up just in time.
• 1:41:54
  And then we've also been as a, as a result of pandemic.
• 1:41:59
  And even before that we working working on Dual sourcing and
• 1:42:04
  triple sourcing some of our critical components.
• 1:42:06
  As a result of it, we have suppliers around 45 countries in
• 1:42:10
  the world. Are our approach really has been
• 1:42:12
  to bring the manufacturing of subcomponents closer to the
• 1:42:15
  point of consumption. It makes the supply.
• 1:42:17
  Hm, more grain because you're not burning diesel moving stuff
• 1:42:19
  around and it also gives us a security of supply of having
• 1:42:22
  multiple sources in case. One Factory gets taken offline
• 1:42:24
  and that's how we got through. Panda Miguel's.
• 1:42:26
  Yep. So last three years were
• 1:42:32
  incredibly hard carne and I used to look lot longer pretty covid.
• 1:42:38
  But on a serious note, pandemic really changed the game of
• 1:42:43
  Supply Chain management. We were just not being a regular
• 1:42:47
  supply chain person. We have rolling up our sleeves
• 1:42:50
  and negotiating with state governments city managers
• 1:42:55
  mayor's to help kick-start our suppliers to restart their
• 1:42:59
  operations. You know, there are some really
• 1:43:02
  inspiring stories where our supply chain team really fought
• 1:43:07
  hard to make sure. Sure, that the factories don't
• 1:43:10
  don't go down. And it's also a testament to the
• 1:43:14
  resilience of our supplier Partners.
• 1:43:16
  Having said that, you know, it was not all about resilience and
• 1:43:22
  never give up attitude. We had also laid the foundation
• 1:43:26
  of having dual sources and also, you know, in many instances, we
• 1:43:33
  had many, you know, things that that the that we had simplified,
• 1:43:42
  the, the parts that we had simplified, the supply chain
• 1:43:45
  design that we had simplified to, to wait.
• 1:43:47
  We were able to Get to the pandemic in much better shape
• 1:43:51
  than the rest of the automotive it first has to adapt.
• 1:43:54
  You know, we encountered issues one at a time, it was really a
• 1:43:57
  team effort. We heavily relied on the
• 1:43:59
  government relations team. You literally had to the code
• 1:44:03
  pandemic, as you're aware, you know, happened in different
• 1:44:05
  countries at different times, but unfortunately, you need all
• 1:44:08
  the parts to make the car. So at any given point, there was
• 1:44:10
  one Factory that was lying down the road.
• 1:44:12
  One jurisdiction that was being, you know, that didn't want to,
• 1:44:15
  they want to shut down the factory because they hadn't had
• 1:44:16
  an outbreak. She's learning.
• 1:44:18
  The rules understanding coming up with argument points.
• 1:44:20
  Leveraging the team and getting those.
• 1:44:23
  It was a, it was a difficult but good experience.
• 1:44:25
  Actually, we learned a lot. Yeah.
• 1:44:26
  And as we were just figuring out to work with covid, then the
• 1:44:30
  port congestion hit in on the west coast Logistics.
• 1:44:33
  Lead time went up by, you know, twice the twice, the amount and
• 1:44:37
  we were operating with less than 35 percent accuracy.
• 1:44:41
  In the system that means at any given point.
• 1:44:43
  The ETA is of the vessels were 75% off.
• 1:44:48
  So we figure out how to work with that and then the chip
• 1:44:52
  shortage hits. Yeah.
• 1:44:54
  The chip shortage. That was a nightmare, but we got
• 1:44:58
  through it. I mean, you know, I think it's a
• 1:45:01
  testament to their tenacity of the team and the resilience of
• 1:45:03
  our supplier Partners, it's mostly behind us.
• 1:45:06
  And I think the results speak for themselves, Tesla was able
• 1:45:10
  to grow production volumes at a time where the rest of the
• 1:45:13
  industry had a hard time, staying flat.
• 1:45:15
  I think they shrunk actually cumulatively And, you know,
• 1:45:18
  we're often asked the question, how do we manage through it?
• 1:45:21
  And the answer is very much kind of baked through the
• 1:45:23
  presentation today. First, we design our own
• 1:45:25
  Electronics. So, in order to control cost, we
• 1:45:28
  really had to go and build relationships with all the tier,
• 1:45:31
  two suppliers to negotiate them directly.
• 1:45:34
  So these relationships are formed, you know, over ten years
• 1:45:37
  ago and and we've really gotten to know them.
• 1:45:39
  Well, so when a shortage hit we knew you know who to escalate to
• 1:45:43
  who to call who to ask for help and and really kind of swarming.
• 1:45:47
  The issue at the A point of constrain.
• 1:45:49
  So the best team, you know, strong relationships with
• 1:45:51
  vendors and then we also have a pretty sophisticated capacity
• 1:45:56
  planning function. Tier 2 capacity in each to come
• 1:45:59
  on earlier than Tier 1 capacity. Tier 3 capacity has to be even
• 1:46:02
  earlier and our team does a very good job of having that
• 1:46:05
  strategic alignment, with all our partners to make sure that
• 1:46:08
  the capacities ready before we need it.
• 1:46:11
  We're a mission based company. Our goal is to put as many V's
• 1:46:13
  out there. We did not let the pandemic
• 1:46:15
  phases, we were full throttle and it was pedal to the metal in
• 1:46:18
  terms of her demand signal. That also helped the, the
• 1:46:21
  Silicon industry has very long lead times.
• 1:46:23
  They invest billions of dollars and sophisticated Fabs that take
• 1:46:27
  years to build. So, that stability and demand is
• 1:46:30
  something they appreciated and it's something that is very,
• 1:46:33
  very much a Tesla thing. So, this was looking back,
• 1:46:39
  right? This is how we got here.
• 1:46:40
  This is how we got from 0 to 40 thousand cars a week.
• 1:46:43
  Let's talk about how we get to 20 million.
• 1:46:46
  So the question gets asked, you know, it's Tesla going to
• 1:46:48
  basically break the chip industry.
• 1:46:50
  If he grow to 20 million Vehicles, the answer is no.
• 1:46:53
  But let me walk you through a logic on that.
• 1:46:55
  So as it stands right now a Tesla enabled with the full
• 1:46:59
  self-driving, computer has four times the Silicon content of a
• 1:47:03
  regular vehicle. So at about the 2 million run
• 1:47:06
  rate, we're consuming about 700 thousand, twelve inch wafer
• 1:47:10
  equivalents. That's the industry term that's
• 1:47:11
  thrown around for capacity and silicon.
• 1:47:14
  That's how much we're using. So little bit below a million
• 1:47:16
  Ian and the global wafer capacity, based on the estimates
• 1:47:18
  we've seen is about 135 million. So we're about half a percent of
• 1:47:23
  the industry, not a lot. Now, if you get to 20 million,
• 1:47:26
  when we get to 20 million, and if we don't make any
• 1:47:29
  simplifications in terms of part count, which is also a separate
• 1:47:33
  work stream, I think Colin talked about it, we talked about
• 1:47:35
  it, we're not going to have all these different components for
• 1:47:37
  really going to simplify the architecture because simpler is
• 1:47:40
  better, but even if you don't do that and you take the most
• 1:47:43
  pessimistic case, we're going to need eight million wafers.
• 1:47:46
  Okay. And the industry is going to
• 1:47:47
  scale the 20 to 200 million by that point.
• 1:47:50
  So still we're less than five percent of market share, and if
• 1:47:53
  we run the same offense, we ran all along and continue to work
• 1:47:56
  with our partners and, you know, invest in Fabs and invest in
• 1:47:59
  capacity, will have the capacity ready by the time, we need it.
• 1:48:02
  So I really don't see this as an impediment, from a supply chain
• 1:48:07
  perspective to get to 20 million.
• 1:48:13
  Now, I would like to talk about a page that we took out of the
• 1:48:16
  electronic Playbook and applied it to the mechanical side.
• 1:48:21
  Our vehicles use currently the heat pump base technology for
• 1:48:24
  the thermal system but before heat pump our vehicles used to
• 1:48:28
  have dispersed thermal system connected by hoses.
• 1:48:32
  Engineering was inspired from Circa circuit boards, and then
• 1:48:37
  created a system, which was extremely integrated, where 60%
• 1:48:43
  of the system was the size of a basketball side basketball where
• 1:48:47
  it contained about 100 components 50.
• 1:48:51
  Ceiling interfaces and several several different manufacturing
• 1:48:55
  processes all with packaged into this tight tight sighs.
• 1:49:00
  Now, why are we talking about this?
• 1:49:02
  Well we launched this system right?
• 1:49:05
  During the outbreak of the pandemic in model.
• 1:49:07
  Why? The team was forced to manage
• 1:49:11
  100-plus components from different suppliers around the
• 1:49:14
  world remotely. So, therefore videos playing at
• 1:49:21
  the same time, one is the video of the manual line that was
• 1:49:25
  created. When we started the production
• 1:49:28
  and this manual line produced, a lot of quality issues, let alone
• 1:49:33
  throughput. So we quickly you know, pivot it
• 1:49:37
  to a semi-automated line. But in spite of that, although
• 1:49:41
  the three throughput was great, the quality issues sustained,
• 1:49:48
  So, our team decided that fully automated is the only way to go.
• 1:49:54
  So before we automate the line, we created a 3D simulation.
• 1:50:00
  Now, not many companies or teams have this capability.
• 1:50:04
  We're at this point. Now when a part gets pecked, we
• 1:50:09
  have the ability to create a 3D simulation of all the complex
• 1:50:13
  subassemblies that would be involved in making the part and
• 1:50:17
  tell the supplies exactly what equipment to purchase what to
• 1:50:21
  test and an end. It an end-to-end line in
• 1:50:24
  simulation. And that's what we did here.
• 1:50:27
  So, with the help of a simulation, we had a very
• 1:50:30
  detailed sub-assembly cell, which was juxtaposed with the
• 1:50:36
  with the suppliers Factory layout.
• 1:50:39
  And then we after a back-and-forth of iteration and
• 1:50:45
  optimizing the processes, we implemented this fully automated
• 1:50:49
  line in less than eight months. And the results were remarkable.
• 1:50:57
  99% reduction in labor force. So in a line that would need
• 1:51:03
  thousand Associates to build the super manifold in full volume.
• 1:51:08
  Now just takes 10, 10, Associates, And the quality also
• 1:51:13
  went up by an order of magnitude.
• 1:51:16
  We have a defect rate of less than 0.005 percent and a super
• 1:51:22
  manifold rolls out every 7 seconds because we created many
• 1:51:26
  automated lines and these factories are the size of two
• 1:51:30
  football fields, by the way, Russian at the one, I play on
• 1:51:34
  that, 99% two football fields. So, why are we talking about
• 1:51:40
  this? In a world where labor costs are
• 1:51:44
  increasing, labor participation is decreasing, there is a huge
• 1:51:48
  turnover and the cost of implementing automation is going
• 1:51:53
  down by 3x in the next 10 years. This is how we are thinking
• 1:51:58
  about complex assemblies and we are going to push this type of
• 1:52:02
  thinking into our supply base more and more who will be
• 1:52:05
  building our complex assemblies. This is the only way we think we
• 1:52:09
  can confidently scale to the Auntie million Vehicles Target.
• 1:52:17
  To summarize. We have laid the foundation in
• 1:52:20
  getting ready to execute master plan 3.
• 1:52:24
  Our strategy is supply chain design simplification.
• 1:52:29
  We are going to make sure that we have more control into the
• 1:52:32
  tears of the supply chain. We are going to grow responsibly
• 1:52:37
  and sustainably with our long-term partners and will
• 1:52:41
  automate. Yeah.
• 1:52:42
  It's right. You know going from 0 to 40
• 1:52:44
  thousand cars a week was was tough.
• 1:52:47
  I mean it took, it took a lot of Trial and error.
• 1:52:49
  It took a lot of learning, there was a lot of pain.
• 1:52:51
  There's a lot of mental Scar Tissue as you on calls it
• 1:52:54
  through that process. But now, that we're at that
• 1:52:57
  level and the foundation is set going from 40 to 400 is not
• 1:53:00
  going to. It doesn't really faze us.
• 1:53:02
  We have a capable team. We've got capable external
• 1:53:05
  partners that have, you know, gone through hell with us
• 1:53:07
  leadership. Team has galvanized behind the
• 1:53:09
  mission and we're going to get it done.
• 1:53:12
  So I'd like to thank you for your time and hand it off to
• 1:53:14
  Drew and Tom for the next presentation.
• 1:53:17
  Thank you. Hi everyone.
• 1:53:27
  My name is Tom Drew. I'm now responsible for Global
• 1:53:30
  Production as cells and deliverance service.
• 1:53:32
  I joined the company back in 2014 and I've been running the
• 1:53:36
  company's business in Korea. China, and a pack today, I'm
• 1:53:40
  representing all the Giga factories and talk about how we
• 1:53:43
  can make more cars and faster. So right now, we have four
• 1:53:50
  vehicle factories on three continents, to serve our
• 1:53:53
  markets, the total number of manufacturing employee is over
• 1:53:57
  65,000, this is also including our amazing team in Reno.
• 1:54:01
  There are making powertrains battery packs or drive unit for
• 1:54:04
  a vehicle factories. The total installed capacity in
• 1:54:07
  all these vehicle factories about two million cars a year,
• 1:54:10
  that's it. We always looking for
• 1:54:11
  opportunity to grow more capacity from the existing
• 1:54:14
  footprint. So, you expect this number will
• 1:54:17
  grow Time. To build more factories is the
• 1:54:24
  start to build a more cars and we certainly expert of it.
• 1:54:28
  This is the before, and after image showing you the progress,
• 1:54:32
  we made back in 2019 when we build a big Factory change in
• 1:54:35
  Shanghai. So in nine and a half months
• 1:54:38
  time we built a Greenfield project to against established a
• 1:54:43
  vehicle Factory in the three months later, we start deliver
• 1:54:46
  our first customer vehicle people.
• 1:54:49
  Are we asking? You know how Tesla can build a
• 1:54:51
  factory that fast. Really we learned from our
• 1:54:56
  Fremont Factory a lot, we talked to the Survivor from the
• 1:55:00
  production Hill and try to avoid all the mistakes we made and we
• 1:55:05
  decided to design a straight line with the minimum number of
• 1:55:08
  up and downs and turns for easier manufacturing ability and
• 1:55:11
  either construction, we also challenge the all the
• 1:55:13
  assumptions that people ever know to build a vehicle Factory,
• 1:55:17
  when we delete in the simplified all the redundant, He's in a
• 1:55:21
  buffers that's helped us to save a lot of time.
• 1:55:24
  Also, we have a very strong in house construction team in Tesla
• 1:55:31
  if there ever a job, cannot be done by others better and we
• 1:55:35
  bring this in health so we have this in the house.
• 1:55:37
  Construction team have a full control over all the activities
• 1:55:41
  on site from design, engineering to construction site management.
• 1:55:46
  So this team not didn't just build bigger Shanghai.
• 1:55:50
  They also Built bigger, Berlin, Giga taxes, and getting them out
• 1:55:53
  of. And they're really the hero
• 1:55:55
  behind the scenes, Like Larson shared earlier going forward
• 1:56:01
  with the new platform, the more modular design and density,
• 1:56:06
  intensified gigafactory will be able to make more cars from
• 1:56:10
  evenly reduce the footprint that also means we can build more
• 1:56:14
  Factories at the same time. With the drawing, a figure of
• 1:56:17
  all manufacturing teams around the globe.
• 1:56:22
  I'm happy to share early this morning.
• 1:56:24
  Will hit the four million Mark for total Tesla ever built.
• 1:56:32
  And they are the formulas vehicle actually built in this
• 1:56:36
  Factory when you took a tour on our shop floor, you probably
• 1:56:40
  work past it. All right, so true because 12
• 1:56:43
  years to build the first million vehicle and about 18 months to
• 1:56:50
  the second building. The second million, the third
• 1:56:52
  million, took us 11 months, and just shortly.
• 1:56:55
  Lesson is 7 months. Would build the four million
• 1:56:57
  cars. So we're getting better at a
• 1:57:00
  faster, exponential growth. Really kudos to the team.
• 1:57:06
  So what it takes to run a giga Factory?
• 1:57:10
  Well, if you have a 600 robots, 10,000 train employees or 5,000
• 1:57:15
  human than 5,000 Optimist and hundreds of process.
• 1:57:19
  If you can do it sounds simple but it's extremely hard.
• 1:57:23
  So there's two key metrics as we plumped.
• 1:57:25
  Predominantly focused on it is an overall equipment
• 1:57:29
  Effectiveness and the cycle time in Tesla we're setting the
• 1:57:33
  passing grade for our vehicle factories with D %.
• 1:57:36
  OE, + 45 second cycle time. What does means the oee?
• 1:57:42
  Really evaluate the equipment up time the machine performance and
• 1:57:47
  quality. Simply put this is the actual
• 1:57:51
  production time on a good quality product versus the plan.
• 1:57:56
  The productive time, the higher the better 45 second cycle time.
• 1:58:02
  That means you know you expect every 45 seconds, it doesn't car
• 1:58:06
  running off the final assembly line in the factory and the
• 1:58:10
  faster we rent. The faster we can get the
• 1:58:13
  economy of scale. If you look at the chart on the
• 1:58:16
  right on the left, Shanghai be able to significantly drop our
• 1:58:20
  labor hours per car during the ramp.
• 1:58:24
  The little dip. There is a happened in the last
• 1:58:26
  cue to 2022 because of the covid shutdown and And on the right of
• 1:58:32
  the Fremont model, why shop, even?
• 1:58:35
  This is a six year old facility. The team they're still be able
• 1:58:38
  to optimize the material flow eliminate all the single point
• 1:58:41
  of failure and a drive, higher output, hence reduce the labor
• 1:58:46
  hours and actually this Factory keep sending new record
• 1:58:51
  yesterday. They just had a new Factory
• 1:58:53
  Daily Record. Congrats Fremont team.
• 1:58:57
  Not really. So, how can we reduce the cycle
• 1:59:05
  time? Keep improving efficiency.
• 1:59:08
  We follow the philosophy along shared about the building
• 1:59:12
  rocket, which is questioning, basically found the right
• 1:59:15
  problem to solve. And we start with delete
• 1:59:19
  simplify. Then we try to accelerate
• 1:59:22
  pressurize, the line and find the if the solution actually
• 1:59:25
  work and at last, we think about it automation.
• 1:59:29
  There's one example in Gig. Factory Shanghai in the paint
• 1:59:33
  shop. We find that there is a overlap
• 1:59:37
  baking range between the PVC sealer and the top coat.
• 1:59:42
  It was done by two different ovens and we decided to combine
• 1:59:45
  the two processes eventually that helped us didn't just help
• 1:59:50
  us to reduce the cycle time, but also save it.
• 1:59:52
  A nine percent of the energy consumption and the nine percent
• 1:59:55
  of the CO2 emission also we common eyes, the snx, it design
• 2:00:01
  in Fremont Right now we have a common body and common headlamp.
• 2:00:07
  We eliminate eliminated about a 40 parts and a reduced about a
• 2:00:11
  10% of the cycle time. So all these improvements really
• 2:00:15
  helped us to get to a faster ramp trajectory.
• 2:00:21
  And we didn't just learn from these high volume and existing
• 2:00:24
  factories. We also learn from the new
• 2:00:26
  factories most recently, our key Berlin Factory team, Men today
• 2:00:32
  5G private Network on the shop floor helped us to overcome
• 2:00:37
  about the 90% of the psyche over cycle issues for a particular
• 2:00:41
  process. And in GA shop we're going to
• 2:00:44
  soon implemented as a globally. So Right now we have an
• 2:00:51
  integrated Global organization from from production all the way
• 2:00:56
  to the cells in telomere service.
• 2:00:58
  This is will help us to strengths feedback loop between
• 2:01:01
  manufacturing and the service. We obviously want to bring a
• 2:01:04
  delight for customer ownership, experience to our car owners and
• 2:01:11
  with this direct feedback loop, we be able to turn customer
• 2:01:15
  excavations and the feedback into a quick actions Improvement
• 2:01:18
  on the shop floor. Immediately with that efforts in
• 2:01:21
  the past six months would be able to reduce the time in
• 2:01:24
  service and early ownership service.
• 2:01:27
  Also service appointment time significantly.
• 2:01:33
  So for factories, not enough, if we want to hit the 20 million
• 2:01:38
  car, a year Target, we're going to keep building a new
• 2:01:41
  factories, New Alliance, and that will come come up with new
• 2:01:46
  products as well. And with all giggle factories,
• 2:01:50
  and now in the one organization would be able to replicate
• 2:01:54
  unified ideas course factories. Also, we will help the new
• 2:01:58
  factories that rent faster and also Produce a better quality
• 2:02:03
  and the lower cost. Now I'm going to hand over to
• 2:02:07
  Julia. Talk about a cell manufacturing.
• 2:02:10
  Thanks, Tom. Yeah, we're not just talking
• 2:02:12
  about new vehicle factories but also new cell factories and and
• 2:02:16
  we're going to talk about how we follow the same model when we do
• 2:02:20
  to do so. But first, I thought I'd just
• 2:02:21
  provide a little bit of an update on cell production.
• 2:02:25
  So remember battery day we showed this video with the spoon
• 2:02:30
  and how we went from dry. Film, let's just say there's no
• 2:02:33
  spoon now. So many of you in this room saw
• 2:02:38
  this on your tour today but you know, here is our dry electrode
• 2:02:41
  machine here in Texas. One of the lines we have
• 2:02:43
  installed here fully automated no spoon from powder and foil
• 2:02:48
  into coated electrode out from from a peak productivity
• 2:02:52
  perspective. Per tool perspective is over 20
• 2:02:55
  times the productivity of the tools that we showed folks on
• 2:02:57
  the tour and Kato back and Battery day.
• 2:03:00
  So we've made a lot of progress on The key, one of the key parts
• 2:03:03
  of the cell manufacturing process.
• 2:03:06
  And we've also continued to focus on refining the way we
• 2:03:10
  make cells and the factories that the cells go inside.
• 2:03:14
  We as you can see on this chart from typical 2170 cylindrical
• 2:03:18
  cells to 4680 made a huge leap, which is basically a 5x
• 2:03:22
  reduction in the factory footprint and volume and volume
• 2:03:27
  and footprint. And then from going, from what
• 2:03:30
  we did in Fremont, and our pilot line, To Texas, we improved
• 2:03:33
  further. And we're proving again, when we
• 2:03:35
  go into Nevada and what this actually represents is a series
• 2:03:39
  of actions taken by a very integrated holistic design team
• 2:03:44
  across, you know, the product design, the manufacturing
• 2:03:47
  design, the process design equipment design in the facility
• 2:03:49
  design. They all need to work together
• 2:03:51
  to make this happen and you can see on the bottom.
• 2:03:54
  Just as an example of simple, Simplicity up investment down in
• 2:03:58
  scaling up, you know, Parts, we've reduced the number of
• 2:04:00
  parts in the cell. Never process is significant
• 2:04:03
  reduction in number of processes and and collaborating with those
• 2:04:07
  five design groups, we've been able to result, you know, build
• 2:04:11
  a end up in a with a factory that's 10 times smaller
• 2:04:14
  volumetrically, which means it's faster to get built.
• 2:04:18
  It's much lower. Capex per gigawatt hour output
• 2:04:21
  and we can go and scale, you know, to our objectives of, you
• 2:04:25
  know, 240 Total terawatt-hours, One Tower water a year of
• 2:04:28
  stationary storage and 20 million Vehicles a year.
• 2:04:33
  With the scalability that is required to achieve our goals.
• 2:04:38
  And we're not just looking at the cell Factor itself, but also
• 2:04:41
  Upstream materials were necessary.
• 2:04:43
  What you see here is a the rendering of the 50 gigawatt
• 2:04:46
  hours, a year Corpus Christi lithium Refinery that we've
• 2:04:49
  already broken ground on here in Texas, the facility will start
• 2:04:54
  commissioning by the end of 2020 three this year.
• 2:04:57
  This is a good example of something where we're basically.
• 2:05:01
  Think about breaking ground and, and, and starting commissioning
• 2:05:05
  within 10 months and with, with actual production, within 12
• 2:05:08
  months, that's the target similar to what we did in
• 2:05:11
  Shanghai. Again are the result of
• 2:05:13
  collaboration and and you know internal execution of
• 2:05:17
  construction in partnership with local communities through this,
• 2:05:21
  the site is 30 minutes from the Corpus Christi ports.
• 2:05:24
  Located directly on rail. The process route, we're taking
• 2:05:27
  is a direct soda. Ash Leach of the input material,
• 2:05:31
  which Means there's no acid roasting.
• 2:05:34
  We don't have any of the sort of waste products associated with
• 2:05:37
  an acid. Roasting step, we're designed to
• 2:05:40
  consume lithium Spa Jeanine, which is a very commonly traded
• 2:05:45
  lithium Rock. But it is flexible to other
• 2:05:47
  feeds from primary and secondary sources.
• 2:05:49
  And similarly, we're working on our Catholic facility here in
• 2:05:52
  Texas. Maybe we've seen it as you're
• 2:05:53
  driving around, which is a 60 gigawatt-hour year, Catholic
• 2:05:57
  facility behind the main building here.
• 2:06:00
  Ten months build time there as All the install, the equipment
• 2:06:03
  being installed for the first line as we speak and will be
• 2:06:05
  commissioning starting next quarter.
• 2:06:07
  So we are doing this where we need to our plan is not to do.
• 2:06:11
  It always. There are lots of confident
• 2:06:12
  companies out there but we are also trying to sort of
• 2:06:15
  accelerate the pace of the industry by trying.
• 2:06:18
  Some new things that are little bit more scalable and and
• 2:06:22
  de-risking certain innovations that improve productivity per in
• 2:06:25
  terms of capex per gigawatt hour and things like that.
• 2:06:29
  And with that, I think that sums up our efforts.
• 2:06:31
  On the factory side. Mike you want to come up and
• 2:06:34
  we'll talk about energy, right? All right, take it away.
• 2:06:46
  All right here, you can. Hello everyone.
• 2:06:49
  My name is Mike Schneider, I lead our mega pack organization.
• 2:06:53
  I've been with Tesla for almost nine years.
• 2:06:56
  I'm excited to talk to you today about the mega pack, the
• 2:06:58
  product, and the business, and some of the exciting things we
• 2:07:00
  have ahead of us. From the beginning with this
• 2:07:04
  business, we have always focused on building successful projects
• 2:07:09
  for our customers and not just the batteries.
• 2:07:12
  In the Box, we've built a hardware and a software platform
• 2:07:16
  that is able to adapt to environments all over the world.
• 2:07:20
  And is able to scale for from small island projects to large
• 2:07:24
  gigawatt-hour, scale batteries. We've invested our time and our
• 2:07:28
  talent to best understand every aspect every step in the process
• 2:07:34
  and every risk of a project to ultimately provide a solution
• 2:07:38
  that is that is thoughtful intuitive provides.
• 2:07:41
  Great Value to our customers and is as plug-and-play as possible.
• 2:07:47
  And yeah, it just looking at these projects.
• 2:07:49
  It's incredible to see them all. Come together in one in one
• 2:07:52
  Montage. Being the future.
• 2:07:54
  Yeah. Compressed.
• 2:07:55
  It's 10 years of building these projects and and it's incredible
• 2:07:58
  to see the impact and to see what we have in front of us.
• 2:08:01
  Absolutely. We're on our sixth generation of
• 2:08:05
  our industrial product, which is the mega pack Excel that were
• 2:08:08
  building out of the out of Lathrop California.
• 2:08:11
  We've deployed over 16 gigawatt hours of industrial and
• 2:08:14
  residential products across 50 countries and really mega pack
• 2:08:17
  is a market leader. Its Best in Class in efficiency,
• 2:08:21
  reliability energy density and easiest to install lowest cost
• 2:08:26
  to install. And we call it Excel because it
• 2:08:29
  is actually the, like, largest heaviest object that you can
• 2:08:34
  transfer around the roads of the world.
• 2:08:36
  Without, you know, having to shut them down and get crazy
• 2:08:38
  permits and things like this. It is extra large, there is not
• 2:08:42
  going to be an XXL. Let's just say it that way.
• 2:08:47
  There's been incredible. Demand for the project are for
• 2:08:49
  the product. 2023 is going to be a great year.
• 2:08:51
  We have gigawatt-hour scale projects being built.
• 2:08:54
  The laser Factory continues to scale and ramp up and those new
• 2:08:57
  products on the near-term roadmap.
• 2:08:58
  So a lot to look forward to How did we get here?
• 2:09:03
  Number one, a maniacal, focus on all aspects of delivering
• 2:09:06
  stationary storage. What do we mean by that?
• 2:09:09
  What are some examples, as you see here, the mega pack
• 2:09:12
  enclosure for one example, we designed the wire ways directly
• 2:09:16
  into the base of the enclosure. So typically you need to install
• 2:09:18
  a bunch of conduits and cables under the ground.
• 2:09:21
  Now you can lift them up and install them directly into the
• 2:09:23
  base lower increasing the speed of deployment and lowering the
• 2:09:27
  cost. Next, we see the batteries in
• 2:09:29
  the power electronics. And the Power Electronics is
• 2:09:32
  really the Beauty and the Elegance of the mega pack.
• 2:09:35
  So, this is what converts the DC power to AC power on the grid.
• 2:09:39
  The brain is built directly into the product.
• 2:09:42
  It is able to connect to any Grid in the world right out of
• 2:09:45
  the box. It allows it to all the mega
• 2:09:49
  packs on a given site to work as one unit.
• 2:09:53
  And really it's it's what makes it what it is.
• 2:09:55
  I think it's the most incredible feature of the mega pack.
• 2:09:58
  As you build out the full site. What we end up with is the most
• 2:10:02
  energy dense Solution on the market upwards of 300 megawatt
• 2:10:05
  hours per acre. And just as a frame of
• 2:10:07
  reference, this solution is two times more power dense than a
• 2:10:11
  typical gas peaker plant. So this is the future, this this
• 2:10:15
  this is where we're going. This is the product that retires
• 2:10:19
  the fossil fuels. Yep, one power plants at a time.
• 2:10:23
  We talk about Power Electronics, a little bit more Tesla, is a
• 2:10:26
  leader in Power Electronics. We've deployed over one point
• 2:10:29
  for TW across energy storage and vehicle and we deliver more
• 2:10:34
  power Electronics than the solar and the wind industry combined
• 2:10:37
  on a per annum basis and Power Electronics.
• 2:10:40
  It's hard to under to overstate how impactful they are.
• 2:10:44
  They are really the glue in the sustainable energy economy
• 2:10:47
  between Generation storage and the end-use, you know, the,
• 2:10:52
  those Power Electronics devices are switching, you know,
• 2:10:54
  thousands of times per second, hundreds of thousands of times
• 2:10:56
  per second to efficiently, react to whatever is needed either in
• 2:11:01
  the car and the grid and because they are so like sort of
• 2:11:05
  software driven at their core. They can provide functionality
• 2:11:08
  that hasn't been available to the Grid in the past and it's
• 2:11:11
  one of the reasons why Renewables and storage together
• 2:11:13
  are such a great solution. Yeah.
• 2:11:17
  And one reason that we focus on Power Electronics and control so
• 2:11:21
  much is because of the impact that it has on the projects
• 2:11:24
  directly. So just as an example, here on
• 2:11:27
  the left, we have it, we had a, have a firmware feature that we
• 2:11:30
  call virtual machine mode. And what virtual machine mode
• 2:11:34
  mode is, is it contributes to grid stability?
• 2:11:37
  Like a car shock absorber that dampens oscillations or
• 2:11:40
  vibrations and keeps the ride smooth.
• 2:11:43
  So you can imagine if you can't, if you don't dampen those, those
• 2:11:46
  vibrations, the vehicle could lose control.
• 2:11:48
  Or or someone could get hurt on the grid.
• 2:11:52
  You could have a blackout. We have one grid operator,
• 2:11:55
  that's utilizing virtual machine mode and said they will not
• 2:11:57
  operate their Grid at 100% Renewables.
• 2:11:59
  Unless they have this feature unless they have virtual machine
• 2:12:01
  mode working. Yeah.
• 2:12:03
  And they tried and it didn't work very well and virtually
• 2:12:05
  what why is it virtual machine? But it's like synthetic inertia
• 2:12:07
  where you turn the battery power plant into like through software
• 2:12:11
  behaving, like it's a giant spinning machine literally and
• 2:12:15
  and and and that like Inertia stabilizes the grid and you
• 2:12:20
  don't but you don't need a giant spinning machine.
• 2:12:22
  You don't need like a huge fossil fuel power, plant or a
• 2:12:24
  giant Hydro turbine, you just have the battery to it and you
• 2:12:26
  can program it to whatever you need at that part of the Grid or
• 2:12:29
  even have it, be dynamically changing as the the grid
• 2:12:31
  conditions change, right? That's why we focus so much on.
• 2:12:34
  And we believe it's the future of how batteries are going to
• 2:12:36
  add value is going to be more about adding power stability.
• 2:12:38
  Just as it is about energy. Yep, Auto bitter is another
• 2:12:42
  software feature. I wanted to mention Auto bitter
• 2:12:44
  is an autonomous energy trading platform that in its most basic
• 2:12:47
  Basic senses, buying energy low and selling it high and the the
• 2:12:52
  owner can net the difference of that.
• 2:12:53
  But operating battery storage is actually very complex.
• 2:12:58
  Upwards needing upwards of hundreds or even thousands of
• 2:13:01
  decisions. Every 5 minutes.
• 2:13:02
  It's much more complex than a PV, a solar plant or a wind
• 2:13:06
  plant or a thermal thermal battery Earth thermal Generation
• 2:13:10
  Plant. And it's mostly because of the
• 2:13:12
  versatility and the ability of the battery to provide so much
• 2:13:15
  value in different ways that these decisions need to be made
• 2:13:17
  in real time in order to optimize and get the most value
• 2:13:20
  out of the batteries that can. So there wasn't a solution that
• 2:13:23
  existed in the market. So we built one ourselves in the
• 2:13:26
  market that we've been to markets.
• 2:13:28
  We've deployed them in and Australia in Texas.
• 2:13:30
  In the UK, it is proven to be a market leader.
• 2:13:32
  We're continuing to invest in Auto bitter as we expand to New
• 2:13:35
  Markets. Second point of how we got here,
• 2:13:40
  a Relentless focus on speed of execution.
• 2:13:43
  And, and there's two points to be made here.
• 2:13:45
  First. One is in building factories.
• 2:13:47
  What you're seeing here is a time lapse of the mega Factory.
• 2:13:50
  That was built in Lathrop. We took it from a JC Penney
• 2:13:53
  Distribution Center to a world-class manufacturing
• 2:13:56
  facility and less than 12 months.
• 2:13:58
  Which is, which is incredible. And really the way that we did
• 2:14:01
  this is similar to what my colleagues have described.
• 2:14:03
  It's leveraging the vertical integration of Tesla.
• 2:14:07
  It's getting the vehicle manufacturing team in the same
• 2:14:10
  room as construction and engineering, and making
• 2:14:12
  decisions quickly with all the decision, with all the
• 2:14:14
  stakeholders and decision-makers.
• 2:14:17
  We're using the lessons learned from Lathrop as we speak as we
• 2:14:19
  plan for our next factories. Second, it's about installing
• 2:14:25
  projects faster. I've mentioned this already but
• 2:14:28
  Plug and Play is kind of at the core of what we're going for
• 2:14:31
  over the past four years. We've we've increased the
• 2:14:34
  installation speed by 4X and we've reduced the total labor
• 2:14:39
  involved, across both construction and Manufacturing
• 2:14:42
  by 3x. We think that this is key to
• 2:14:45
  unlocking, our ability to scale. And our customers ability to
• 2:14:48
  scale with us. We need to make be laser focused
• 2:14:50
  on reducing that time from when the Mega pack leaves the factory
• 2:14:54
  to 1 it is operational on the grid.
• 2:15:00
  And it's not just about centralized storage.
• 2:15:03
  It's also about distributed storage the future roadmap of.
• 2:15:07
  Well, at least here at Tesla. I'm and and maybe some of, you
• 2:15:10
  know, about this sort of retail plan, we have here in Texas and
• 2:15:14
  I just wanted to talk a little bit about how that all fits
• 2:15:16
  together. So Tesla electric, really
• 2:15:19
  unlocks the full value of distributed energy and storage
• 2:15:22
  products. So it enables, our customers to
• 2:15:26
  become their own utility. The data on the page is from
• 2:15:29
  South Australia, virtual power plant, another Tesla electric
• 2:15:34
  set up that we have over the over 20 22.
• 2:15:38
  It's 5,000 customers that we have the data for you can get a
• 2:15:41
  sense for what's happening. So just if they were being
• 2:15:44
  provided to fault utility service and you looked at the
• 2:15:46
  cost of serving them that electricity on average hundred
• 2:15:49
  forty dollars a month if they all had solar and powerwall but
• 2:15:52
  the solar Empower all wasn't interacting with the grid wasn't
• 2:15:54
  participating in the energy Market.
• 2:15:56
  The cost of serving that customer would go from a would
• 2:15:58
  have Go from $100 a month to $70 a month.
• 2:16:02
  If Tesla electric is operating. Those assets in an intelligent
• 2:16:06
  way to benefit the grid, Tess electric being basically a
• 2:16:09
  software that we developed out of Auto bidder for the purposes
• 2:16:12
  of of these distributed energy resources, we can actually pay
• 2:16:18
  the customers to bring their their, their Energy Services to
• 2:16:21
  the grid. That's what happened in
• 2:16:22
  Australia last year. And it's always a little bit of
• 2:16:25
  a special case there. They represent the future, South
• 2:16:28
  Australia in particular Solar and wind, supplied, 70% of South
• 2:16:31
  Australia's energy in 2022, that compares to 30% in Texas, and
• 2:16:36
  35% in California, but it but this is a seat.
• 2:16:39
  This is an indication of where this is all headed of both
• 2:16:43
  centralized and distributed storage resources.
• 2:16:47
  Providing the the key to unlocking fully renewable grids
• 2:16:52
  and Tess electric is the retail plan that were using to bring
• 2:16:55
  that to bear for our customers. That have our products And you
• 2:17:01
  know, this is how we're rolling it out.
• 2:17:04
  First, as it exists, it's available for in Texas, for
• 2:17:09
  people who have power walls in their homes, but we need to
• 2:17:12
  extend beyond that. Over a billion people live in
• 2:17:14
  markets with competitive retail electricity and there's over two
• 2:17:19
  trillion in annual energy spend. In these markets, this is a huge
• 2:17:22
  opportunity. And the, Our intention is to
• 2:17:26
  bring this to, you know, Market by market in the same way that
• 2:17:28
  we've approached Tesla Insurance where we can In value to our
• 2:17:31
  customers, to reduce their total cost of ownership of our
• 2:17:34
  products. So actually what we're going to
• 2:17:38
  do next and this is pretty exciting here in Texas.
• 2:17:40
  Yeah. By this summer we're going to
• 2:17:42
  offer a retail electricity plan, to people who have our cars
• 2:17:46
  where they can have unlimited overnight home, charging for $30
• 2:17:49
  a month. This is part of reducing the
• 2:17:52
  total cost of ownership of our vehicles.
• 2:17:54
  And the reason why we can do this is because Texas has a ton
• 2:17:57
  of wind and in Texas, the wind blows at night.
• 2:18:00
  So actually, serving these customers with electricity at
• 2:18:02
  night for their cars, is the best thing to do for everyone.
• 2:18:06
  One. That's this is a way to
• 2:18:07
  incentivize people to charge it home at night directly from
• 2:18:10
  Renewable. Power is part of the grand
• 2:18:12
  master plan, we talked about at the beginning.
• 2:18:14
  So we're very excited about this and we do see this similar to
• 2:18:18
  Tesla insurance, as I said as further reducing the total cost
• 2:18:21
  of ownership of electric vehicles.
• 2:18:25
  So putting this all together, we're really at the beginning of
• 2:18:28
  this massive ramp in energy storage deployment.
• 2:18:32
  And yeah, yeah reflecting back on the master plan.
• 2:18:35
  We talked about Of terawatt hour, need intensive, terawatt
• 2:18:38
  hours of stationary storage. So we have our sights set on
• 2:18:41
  annual production rates of one, terawatt hour which is 25 x.
• 2:18:43
  Our capacity at Lathrop in the in the near term.
• 2:18:47
  We see we see really strong demand for the mega pack product
• 2:18:50
  over 100 Giga Watt hours per year in 2023 and growing by over
• 2:18:54
  100 kilowatt hours per year over the next few years.
• 2:18:57
  So the demand will is there and naturally as we continue to
• 2:19:00
  focus on cost and speed and value, the things that we've
• 2:19:03
  mentioned it's clear. We need to build more capacity
• 2:19:05
  and we need to ramp A bit quickly.
• 2:19:07
  So while the challenge is Big, it's also a huge opportunity for
• 2:19:10
  Tesla's, huge opportunity for for the mega pack business, and
• 2:19:13
  I'm excited to see the impact that it's going to make on our
• 2:19:16
  good transition as a my yeah. Alright.
• 2:19:18
  Alright, thank you. Thank you all.
• 2:19:20
  Yes. All right and Laurie Brandon,
• 2:19:29
  come on up. Hi.
• 2:19:36
  I'm Brandon with the legal team. Thank you, not just for coming
• 2:19:41
  but being investors. Now we have we have an active
• 2:19:45
  and engaged board and management team we've met with you.
• 2:19:48
  And today is the culmination of that we've heard you, which is
• 2:19:52
  why we're excited to share this investor date with you.
• 2:19:55
  Now you will hear more from our board at the appropriate time
• 2:19:58
  but for now I'm excited to introduce Lori.
• 2:20:01
  Shelby is going to Gus sustainability and our employees
• 2:20:04
  Lori. Thank you, Brandon.
• 2:20:11
  Hi everyone. My name is Laurie Shelby and I
• 2:20:14
  lead the environmental health safety security and
• 2:20:17
  sustainability team. I've been with Tesla for five
• 2:20:20
  and a half years so I'm going to talk about Teslas impact.
• 2:20:27
  And of course that starts with our awesome team members.
• 2:20:33
  So you've heard a lot today about our amazing products are
• 2:20:37
  process. Our software, you've seen lots
• 2:20:40
  of numbers and lots of graphs, you've also heard a common theme
• 2:20:44
  and that common theme is that Tesla's.
• 2:20:46
  Team is what drives our success and makes a difference.
• 2:20:50
  Whether you're in our factories, and our sales service and
• 2:20:54
  delivery centers are warehouses. We are in the field, you can see
• 2:21:00
  and you can feel the collaboration that's happening
• 2:21:04
  and you also can see the pride that everyone has and working
• 2:21:09
  for Tesla. So before I go on, I want to
• 2:21:13
  give a big shout out and thank you for our Tesla team globally.
• 2:21:17
  So let's That was great. It was great.
• 2:21:27
  Thank you. So we are creating a lot of
• 2:21:31
  meaningful work through our jobs.
• 2:21:34
  We're a huge employer, and we're growing every day.
• 2:21:37
  We're at 129 K and more than half of our team Works in
• 2:21:43
  vehicle manufacturing, and almost 60% of our our team is
• 2:21:49
  based in the US. People want to join our mission
• 2:21:57
  and we need to grow and we need to hire the best.
• 2:22:02
  People Engineers want to work for Tesla.
• 2:22:06
  Look at number one and number two SpaceX and Tesla in 2021.
• 2:22:12
  We got three million applicants, three million.
• 2:22:17
  That's incredible. and our culture, our culture at Tesla is
• 2:22:25
  anyone at any time can raise an idea or make an improvement
• 2:22:31
  suggestion across safety people cost production quality, Because
• 2:22:40
  we have a very simple process called take charge and take
• 2:22:44
  charge as really Employee Engagement and as take charge's
• 2:22:48
  go up as you can see in this graph injuries.
• 2:22:51
  Go down. And you know why people who do
• 2:22:55
  the work, the people who do the work, know how to improve the
• 2:22:58
  work. Sustainability sustainability,
• 2:23:03
  is everything. It's our mission.
• 2:23:04
  You've heard about it all day. Part of our master plan, but at
• 2:23:09
  Tesla, we don't just tick boxes to meet a requirement.
• 2:23:14
  We really focus on the impact that our products and our
• 2:23:18
  software will deliver in 2021. Our customers avoided 8.4
• 2:23:27
  million metric tons of greenhouse gas.
• 2:23:31
  That's equivalent to taking 1.7 million ice Vehicles, off the
• 2:23:38
  road for a full year. Yes.
• 2:23:42
  Yeah we're also focusing on doing really good work and
• 2:23:51
  driving sustainability into our operations.
• 2:23:54
  We have year-over-year reduction in our water, use in our waste
• 2:24:00
  and we're Growing and building a renewable energy program within
• 2:24:05
  our operations. Just like supercharging as you
• 2:24:08
  heard. So it's not just about our
• 2:24:12
  products, it's the way we make our products both phases.
• 2:24:17
  Tesla. Our company is solving both
• 2:24:19
  sides of the clean energy equation.
• 2:24:22
  We've got the clean generation, the clean storage.
• 2:24:26
  We also have zero emission transport as this slide shows
• 2:24:31
  from 2012 to 2020 one. Our products generated more
• 2:24:36
  energy that was consumed by all Tesla vehicles and our
• 2:24:41
  factories, it's amazing. Yeah, I like the claps.
• 2:24:46
  Yeah. Okay.
• 2:24:47
  I'll take all I'll take love. So our vehicles use less
• 2:24:53
  emissions, and gas powered vehicles.
• 2:24:57
  Even in the worst case scenario, 100% coal grid, the model 3 has
• 2:25:05
  less emissions than the average ice car.
• 2:25:09
  And we're only going to get better.
• 2:25:11
  You heard about that today. Every product we make is more
• 2:25:15
  efficient every Factory. We build is more sustainable, so
• 2:25:20
  it's only going to get better and better. so, as we build our
• 2:25:25
  zero-emission future, We've been talking with our stakeholders
• 2:25:31
  with our shareholders about how do we improve our reporting
• 2:25:37
  framework and we are aligning with a t.
• 2:25:41
  C DF? So T cfd.
• 2:25:46
  So we're aligning with the TCF D which I know many of you have
• 2:25:49
  been asking us about. I also want you to stay tuned
• 2:25:53
  because there's going to be much more coming in our impact
• 2:25:57
  report. It.
• 2:25:58
  So thank you all and up next is zachar Court.
• 2:26:13
  Thank you, Lori. All right, so this is our last
• 2:26:19
  section of our prepared remarks. Today will kind of take what
• 2:26:24
  we've discussed today and summarize it at the corporate
• 2:26:27
  level for our financials. Maybe before I jump into the
• 2:26:30
  details here, we've never done an investor day as a company.
• 2:26:34
  We've never brought our leadership team out and ask them
• 2:26:37
  to talk about the things that they're working on.
• 2:26:39
  So you know, I feel very fortunate to be able to work
• 2:26:42
  alongside this group of people and support their teams.
• 2:26:45
  And I also just want to Thank them.
• 2:26:47
  They're in a room in the back listening to this, for all of
• 2:26:49
  the work that they put into today and getting the company to
• 2:26:53
  where it is today. So thank you to the Tesla team.
• 2:27:03
  When we were preparing for today, folks were asking, well,
• 2:27:07
  Zach, what should we talk about? And really the only guidance we
• 2:27:10
  gave folks is well, talk about the things that you're working
• 2:27:13
  on and talk about it in the context of the master plan.
• 2:27:17
  And what did almost everybody talked about today, in, great
• 2:27:20
  detail, all of the work that they're doing to take cost out
• 2:27:24
  because in this industry in this business, You survived or you
• 2:27:29
  die based upon the ability to manage our costs.
• 2:27:32
  And so I'd like to talk about cost as well.
• 2:27:38
  If we look at our longest-running scale production
• 2:27:41
  product, this is the model 3. We reached 5,000 cars per week,
• 2:27:45
  which was our design capacity in mid 2018 and since then we've
• 2:27:51
  taken thirty percent of cost out of this product.
• 2:27:54
  And there's two points I want to make about this.
• 2:27:56
  The first is that cost reduction as you have heard throughout the
• 2:27:59
  course of the day is deeply ingrained in our culture.
• 2:28:03
  And I think one of the most important reasons why we are
• 2:28:05
  here today as Company. The second point that I want to
• 2:28:09
  make is that When we're working on cost reduction, it's easy.
• 2:28:15
  Just to take cost out and make our products worse, but we have
• 2:28:18
  to take cost out and improve our products.
• 2:28:20
  At the same time, this is the hard thing to do but it's the
• 2:28:23
  necessary thing to do to continue to move forward.
• 2:28:27
  And if you look at a version of a model through that we built
• 2:28:29
  today, and you compare it back to a model 3 that we built in
• 2:28:32
  2018, there's a long list of improvements that we put into
• 2:28:35
  the car while also taking costs out of the product, Cost,
• 2:28:42
  reduction, doesn't just come from one place.
• 2:28:44
  There's no Silver Bullet here and we push the boundaries on
• 2:28:48
  volume with the model 3 Program, increasing volume 3 X over this
• 2:28:52
  period of time, we've improved productivity.
• 2:28:54
  So in our Fremont Factory, we are twice as productive.
• 2:28:57
  Now, as we were in mid 2018, we've made a lot of progress on
• 2:29:01
  overhead, efficiencies and product improvements as we've
• 2:29:03
  discussed and a long list of other things including
• 2:29:06
  localization, our Factory in Shanghai.
• 2:29:10
  And as we ramp up volume and as we find efficiencies, we work
• 2:29:13
  with our suppliers to do the same and that leads to material
• 2:29:16
  cost, reductions that improve the affordability of our
• 2:29:18
  products. And as we look forward to our
• 2:29:23
  next gen vehicle, our Target is to reduce 50% of cost.
• 2:29:27
  And we've talked about that a bit earlier today from going
• 2:29:31
  from the model, S NX platform to the 3 and Y platform, we took
• 2:29:34
  out fifty percent of cost. So the task here is to do it
• 2:29:37
  again. And this is very important
• 2:29:40
  because as we improve affordability, The number of
• 2:29:44
  customers who have access to our products dramatically increases,
• 2:29:47
  and as we link this back to our master plan, it enables an
• 2:29:51
  exponential growth in our volume with linear reductions in the
• 2:29:54
  cost of our products. The second point I'll make here
• 2:29:57
  is that in again cost reductions don't come from any single one
• 2:30:02
  place so you can see the buckets here on the vehicle side battery
• 2:30:06
  and powertrain manufacturing cost reductions and others these
• 2:30:09
  buckets are relatively equal in size and so in order to take 50%
• 2:30:13
  Of cost out of the product, we have to go through everything.
• 2:30:20
  but more importantly than just the cost of the car up front,
• 2:30:27
  when we're transitioning to a sustainable economy particularly
• 2:30:30
  with vehicle ownership, it's really important to think about
• 2:30:33
  the lifetime cost of a the products and This chart here is
• 2:30:41
  showing what the total cost of ownership per mile is over the
• 2:30:43
  course of five years. And we have to think about
• 2:30:47
  financing costs insurance cost the cost of power through talked
• 2:30:51
  a little bit about that with the plants that were doing in Texas
• 2:30:54
  wear and tear and maintenance on the cars.
• 2:30:56
  Etc. And we're already at a place
• 2:30:58
  today in the US where a base model 3 on a cost per mile basis
• 2:31:02
  is less than a Toyota Corolla, which is the highest selling car
• 2:31:05
  in the world. And as we move towards our
• 2:31:07
  next-gen platform, we will continue to Reduce this.
• 2:31:10
  And as we work on Robo, taxi variants of this platform, this
• 2:31:15
  cost will come down even further.
• 2:31:17
  And so, this is a product that has that we expect to have
• 2:31:20
  substantially lower cost per mile than the highest volume
• 2:31:23
  products in the world. The next thing on cost
• 2:31:28
  reduction, I want to talk about is operating expenses.
• 2:31:30
  We spend a lot of time on our earnings calls talking about
• 2:31:32
  gross margins and product cost. We don't spend a lot of time
• 2:31:35
  talking about op ex but I think this is one of the really
• 2:31:38
  important parts of our story. So since 2018, we've reduced our
• 2:31:43
  non-gaap Optics. 60%, as a percent of Revenue and 75% on a
• 2:31:48
  per delivered car basis. This is quite staggering, and I
• 2:31:52
  want to go into the details about how we've done this.
• 2:31:55
  The first component of op X which is Hardy, the the single
• 2:31:59
  most important thing here is that we constrain the number of
• 2:32:03
  programs that we work on. So minimizing the number of
• 2:32:06
  programs that were running in parallel.
• 2:32:08
  In the key here is to maximize the revenue and cash generated
• 2:32:12
  from every program that we work on and also maximize the
• 2:32:15
  technology sharing between the programs.
• 2:32:18
  So we talked a little bit about how Power Electronics and our
• 2:32:20
  energy business is shared with Power Electronics in our vehicle
• 2:32:23
  business. If we look at the SGA side of
• 2:32:28
  this, I think the story here is pretty dramatic.
• 2:32:32
  So we're showing here a comparison of sgna per card,
• 2:32:34
  delivered compared to the traditional Auto industry which
• 2:32:37
  is both an oem and the dealership Network.
• 2:32:40
  And our estimates are that were sixty to seventy percent lower
• 2:32:43
  on a per car basis. And if you turn this into
• 2:32:46
  Dollars, this is many thousands of dollars per car.
• 2:32:49
  This is part of the reason why on the last earnings call, I
• 2:32:52
  made the comment that we think about operating margin more.
• 2:32:56
  So, Oh than gross margin as a company.
• 2:32:59
  Because the integration that we have here on the sgna side,
• 2:33:03
  particularly with vertical integration into our dealership
• 2:33:05
  Network. This provides efficiencies that
• 2:33:09
  give us a competitive Advantage from cost structure standpoint
• 2:33:12
  relative to other companies. The other thing that I'll note
• 2:33:15
  here is that this wasn't always the case and it wasn't that long
• 2:33:19
  ago where these bars would be inverted.
• 2:33:22
  And if we go back a little bit in time and we talked about when
• 2:33:25
  we launched model 3, in the Fremont Factory, we went from
• 2:33:29
  two thousand cars per week which was snx adding five thousand a
• 2:33:33
  week to that for model 3, got us to seven thousand cars a week.
• 2:33:36
  You've heard some folks here today, talk about production.
• 2:33:39
  Hell, but I don't think we've ever talked about as a company
• 2:33:42
  as back office operations. Hell Which also happened at the
• 2:33:45
  same time and it's 7,000 cars a week.
• 2:33:48
  Three hundred fifty thousand cars.
• 2:33:50
  A year. We want to go to 20 million cars
• 2:33:52
  a year and we're struggling at 7000 and we're saying to
• 2:33:55
  ourselves. We have to completely rethink
• 2:33:59
  the way that we're managing our back-office operations.
• 2:34:01
  If we have any hope of efficiently scaling this
• 2:34:04
  company, And so this began a process that we're still
• 2:34:08
  continuing to work on today, but it's something that we refer to
• 2:34:10
  as the Tesla operating system. So in the same way that we're
• 2:34:13
  vertically integrated into the software, that manages our cars,
• 2:34:17
  we've also vertically integrated into the software that runs our
• 2:34:19
  company, in all of our major departments in the company are
• 2:34:24
  using, but not all of them yet, but almost all of them are using
• 2:34:28
  our in-house software. Most recently, we removed
• 2:34:31
  third-party recruiting software and we're now running our own
• 2:34:34
  recruiting software as a company.
• 2:34:36
  And this matters a lot because rather than having a complex web
• 2:34:40
  of third-party systems that are generic solution that take a lot
• 2:34:43
  of effort to customize for our particular needs.
• 2:34:45
  We've been able to put in place with an in-house applications,
• 2:34:48
  engineering, team, dedicated lightweight software.
• 2:34:51
  That does only what we needed to do, and nothing more.
• 2:34:55
  And that team works very closely with process Improvement teams
• 2:34:58
  that sit within the business, who are looking through all of
• 2:35:01
  the processes that were managing, as a company and
• 2:35:04
  following the exact same process.
• 2:35:05
  That Tom mentioned in the manufacturing space to look at
• 2:35:09
  our processes. Delete process steps that we
• 2:35:11
  don't need. Simplify process steps that are
• 2:35:13
  existing and automate. What's remaining in the results
• 2:35:17
  of this have been quite, staggering to be frank.
• 2:35:21
  I think it's exceeded our expectations as well.
• 2:35:24
  On the left side of this chart are some examples of the
• 2:35:27
  efficiency improvements that we've seen within our SGA areas
• 2:35:30
  as a result of the strategy, the North American Sales.
• 2:35:33
  Team is four times more efficient now than they were.
• 2:35:36
  In 2018, order, operations, team Financial Services.
• 2:35:40
  I mean these levels of improvement are large and I
• 2:35:43
  don't there's not enough pixels on the slide here to list out
• 2:35:45
  everything, so it's just a couple of examples.
• 2:35:48
  And then in the same way that when we look to take cost out of
• 2:35:51
  our products, we want to feature up the product or improve the
• 2:35:55
  product while taking cost out. The same has been true with the
• 2:35:59
  use of Tesla operating system and that integration with our
• 2:36:01
  process Improvement teams within the company.
• 2:36:04
  And so we've increased East performance and added
• 2:36:07
  capabilities. While also dramatically reducing
• 2:36:10
  cost on a per car basis, there's a list of self-service
• 2:36:13
  functionality that we've put in place, which customers like it.
• 2:36:17
  Also simplifies our operations behind the scenes.
• 2:36:19
  We've used the Tesla operating system to expand into our
• 2:36:22
  captive Insurance face, which we are fully vertically integrated
• 2:36:25
  into, we run our own in-house software.
• 2:36:28
  We have our own agents or owns claim software and, and claims
• 2:36:31
  folks who run our insurance business.
• 2:36:34
  This is also very important because As a technology company
• 2:36:37
  data is very important and we use the systems and the access
• 2:36:42
  to the data in those systems to pull out, extremely granular and
• 2:36:45
  targeted reporting that enables us to see every aspect of our
• 2:36:49
  business in real time. So we can make adjustments to
• 2:36:51
  our operations as needed. And there are sometimes examples
• 2:36:55
  where we say, man, I wish we were tracking this piece of
• 2:36:57
  data. Well, you know, it's a teams
• 2:36:59
  meeting away from the applications engineering team to
• 2:37:02
  make a request and then we can start tracking that data and
• 2:37:04
  make those changes. And so the feedback loop
• 2:37:07
  associated with process Improvement, inside the company
• 2:37:11
  is pretty astounding. The last thing I'll just mention
• 2:37:14
  here which is something I'm personally, quite proud of.
• 2:37:16
  We've made a lot of progress in our clothes processes within the
• 2:37:19
  finance team. Getting our 10-q and 10K filed,
• 2:37:23
  its amongst the fastest of the 20 largest market cap companies.
• 2:37:26
  So great work to the finance team here.
• 2:37:31
  So all of this work on cost reduction is extremely important
• 2:37:35
  because we have a lot of money to spend ahead of us to achieve
• 2:37:38
  our goals within the master plan.
• 2:37:41
  So we've mentioned over the course of the day to day. 20
• 2:37:43
  million annual vehicle production as our Target, one
• 2:37:47
  terawatt hour of annual energy storage production and then
• 2:37:50
  expanding cell production, service and charging in line
• 2:37:53
  with the growth of those other businesses.
• 2:37:55
  And so we've estimated what we think.
• 2:37:57
  The total cost to get, there will be and there certainly are
• 2:38:00
  bars around. Numbers, as we continue to
• 2:38:03
  progress and innovate. But of this, we've spent about
• 2:38:06
  twenty eight billion dollars of that so far in the history of
• 2:38:08
  the company and maybe this total investment looks large.
• 2:38:13
  I actually think it's quite small relative to our Ambitions,
• 2:38:17
  and if you look at our 20 22 operating cash flows and you
• 2:38:21
  just say, well, let's assume some modest growth to that maybe
• 2:38:25
  not all that much. If you're being conservative,
• 2:38:29
  but the ability to pay for this level of investment from the
• 2:38:33
  forecast that we have is very achievable for us.
• 2:38:40
  I'm often asked about Capital allocation, I think our strategy
• 2:38:43
  here is very straightforward. I touched on it a little bit in
• 2:38:46
  our last slide, but just to be explicit about it.
• 2:38:51
  Obviously the priority here is to ensure that we're using
• 2:38:54
  Capital to run the day-to-day operations of the business.
• 2:38:56
  This is a working capital intensive business in this has
• 2:39:00
  to be managed very carefully. And so, you know, over the
• 2:39:03
  course of a quarter, there can be multibillion-dollar swings
• 2:39:06
  and our cash, balance up and down, just depending upon the
• 2:39:09
  timing of when we build and deliver.
• 2:39:10
  Is so we obviously have to have money set aside for that
• 2:39:14
  downside. Protection is another really
• 2:39:15
  important area where we have money set aside for that in over
• 2:39:19
  the last couple of years throughout the pandemic.
• 2:39:21
  I think this just highlights. Why being ready for downtown
• 2:39:24
  side protection is important preserving.
• 2:39:27
  Daily operations is the engine that then generates cash that
• 2:39:31
  allows us to reinvest into the growth of the business.
• 2:39:35
  And I'm very proud that over the last couple of years, despite
• 2:39:38
  the ups and downs in the macro. Environment and interruptions to
• 2:39:42
  supply chains. We have never once pulled back
• 2:39:44
  on our investments in growth, so that remains a huge priority for
• 2:39:47
  us. We do then after investing in
• 2:39:52
  growth, we look at opportunistic ways to spend our cash, these
• 2:39:55
  are more financially oriented where for return exceeds a
• 2:39:57
  hurdle then we can place cash in that.
• 2:40:00
  This is part of our story of cost reduction in the Fremont
• 2:40:03
  Factory. And then on the excess side this
• 2:40:07
  is where the board meets on a regular basis and if thinking
• 2:40:10
  about what our cash flow projections, are looking forward
• 2:40:13
  and whether it makes sense to do BuyBacks or dividends, So to
• 2:40:20
  wrap up the finance portion of the presentation today, couple
• 2:40:25
  of key messages. I just want to leave you all
• 2:40:27
  with here. Is that we use innovation at an
• 2:40:31
  intense level to drive costs down and improve the efficiency
• 2:40:36
  of the business. And the reason this is so
• 2:40:39
  important is it allows us to improve product affordability.
• 2:40:42
  This is particularly the case when we move to our next
• 2:40:45
  Generation platform. Improving affordability, allows
• 2:40:49
  us to comfortably make investments that grow volume,
• 2:40:54
  that volume generates cash that then allows us to make more
• 2:40:57
  Investments. And as we integrate that going
• 2:41:00
  forward, it's our belief as a leadership team here at Tesla
• 2:41:04
  that we're going to achieve unprecedented scale in the
• 2:41:07
  manufacturing space. In this is what's ultimately
• 2:41:11
  required to accelerate the world's transition to
• 2:41:14
  sustainable energy. So with that this completes the
• 2:41:21
  prepared portion of today, we're going to take a short break and
• 2:41:26
  then we'll resume at the live Q&A.
• 2:41:29
  Thank you all very much. Oh, I can talk.
• 2:51:42
  All right. So what was he?
• 2:51:46
  So this is probably the most significant announcement of the
• 2:51:51
  day is that we're excited to announce that we're going to be
• 2:51:53
  building a gigafactory and Mexico.
• 2:52:04
  So, yeah and we'll have a obviously a grand opening and
• 2:52:11
  Groundbreaking and whatnot, but we're excited to announce that
• 2:52:15
  the next Tesla gigafactory will be in Mexico near Monterrey.
• 2:52:21
  So superset. Erotic now do whatever size we
• 2:52:28
  will continue to expand production at all of our
• 2:52:32
  existing factories so including California.
• 2:52:34
  Nevada here in Texas obviously and role in Shanghai.
• 2:52:39
  So we intend to increase. Production at all factories.
• 2:52:43
  So the Giga Mexico would be supplemental to the output of
• 2:52:49
  all the other factories. So this is not to be clear
• 2:52:51
  moving moving output from anywhere to anywhere it is
• 2:52:54
  simply about expanding total Global output.
• 2:52:58
  Yeah so it's gonna be good. So peanut gallery.
• 2:53:09
  All right, so the sewing that's yeah, we're quite excited to
• 2:53:13
  announce that. In fact that they believe the
• 2:53:16
  governor is here. It's hard to see who had
• 2:53:20
  welcomed. Secretary of State, I believe.
• 2:53:30
  So anyway, it's a Look forward to it and having a big event to
• 2:53:36
  for the great groundbreaking and and opening.
• 2:53:38
  So let's see what that we can move to I guess, Q&A we mostly
• 2:53:43
  got significant bench strength here for probably like a bench
• 2:53:50
  and, and we really we should maybe have too many people on
• 2:53:55
  stage, but we'll try to answer questions within reason.
• 2:54:00
  Yeah, we're this is meant to be kind of more of a Long-term sort
• 2:54:04
  of discussion as opposed to, you know, what will be the
• 2:54:09
  production for the rest of the quarter of type of thing.
• 2:54:14
  So let's try to orient our questions towards long-term
• 2:54:18
  value creation and with that far away.
• 2:54:24
  Rob. It's Rod lash for the wolf
• 2:54:31
  research. Very exciting plans about the,
• 2:54:35
  the Next Generation vehicle and powertrains and batteries, I was
• 2:54:40
  hoping can maybe talk to us a little bit about the timeline
• 2:54:43
  for deploying this, and it sounds like it's more than just
• 2:54:47
  a vehicle. This is a kind of a paradigm
• 2:54:48
  change on how vehicles are assembled, how batteries are put
• 2:54:53
  together and everything and does that also, just get reconfigured
• 2:54:58
  into everything that you You so the model lies that are being
• 2:55:02
  built here will be built very differently in the future, maybe
• 2:55:05
  just give us some feel for what happens from here and what's the
• 2:55:08
  timeline for implementation? I'll talk a little bit about
• 2:55:12
  that, but it, broadly speaking, the most profound architectural
• 2:55:20
  changes will be in future Vehicles, retooling.
• 2:55:23
  A factory means bringing the factory down for an extended
• 2:55:26
  period of time and that's what would prefer not to do that.
• 2:55:32
  I think. But we, but there are variants
• 2:55:37
  in how model Y is produced. So we've got a variance where
• 2:55:41
  there's a rear casting where there's a front and rear casting
• 2:55:45
  and we have the structural battery pack and then there are
• 2:55:51
  a number of smaller improvements that occur, but I think for
• 2:55:55
  really, really big changes. Those would be future Vehicles.
• 2:55:59
  Yeah, I don't you guys want ads. Large baby.
• 2:56:06
  So I mean, as far as agree with you 100% Elon, it's really easy
• 2:56:10
  to put Innovations and new vehicles but long-term will
• 2:56:13
  obviously bring them back. We've always talked about that
• 2:56:16
  but we don't want to take our factories down as far as the
• 2:56:18
  timeline goes, you know, we're going to go as fast as we can
• 2:56:21
  left to right. As always you know Ilana alluded
• 2:56:24
  to the fact that Mexico will build our next gen vehicle but
• 2:56:28
  we will also be doing that in our other plants and so it's
• 2:56:31
  really about getting them all up and running week.
• 2:56:34
  Like that to be a huge volume product and we're going to move
• 2:56:38
  that quickly over the next couple of years.
• 2:56:47
  Let's go through them. Thanks.
• 2:56:50
  First, a note, I went to Mexico. My boss that's great.
• 2:56:54
  Congratulations Ilan. Question on applying first
• 2:57:00
  principles thinking and Innovation to an area that up to
• 2:57:05
  now has seemingly been outside of your control and that is on
• 2:57:09
  mining and extraction of some of the key materials.
• 2:57:12
  I believe a couple years ago, you had a patent on sodium
• 2:57:15
  chloride to extract lithium from some from some clays.
• 2:57:19
  And spot. You mean clay and things of that
• 2:57:21
  nature. Any, how does that fit into the
• 2:57:24
  plan of maybe bringing real Innovation into a mining sector?
• 2:57:28
  That could use a little, you know, maybe waking up and
• 2:57:30
  getting those costs down because that could be a real gating
• 2:57:32
  Factor. It seems well, we're going to
• 2:57:36
  address whatever we think the limiting factor is at any point
• 2:57:39
  in time. So we would like to do the least
• 2:57:43
  amount possible. So we don't want to get into the
• 2:57:47
  mining or finding sector. We will do that.
• 2:57:51
  If we have to I do think the the the focus really should be on
• 2:57:57
  refining capacity. You know, do we need to make
• 2:58:03
  still very dry and amount of anode cathode lithium lithium
• 2:58:11
  hydroxide with every book and cop in it.
• 2:58:13
  This is really the refining capacity, that is the biggest
• 2:58:17
  choke point. Yeah, so that's why we're
• 2:58:23
  building a lithium refinery in Corpus Christi.
• 2:58:28
  In terms of the mining companies that are out there and looking
• 2:58:33
  at that part of the value chain. So we do have large suppliers of
• 2:58:38
  lithium right now and we they are aware of, you know, how
• 2:58:42
  we're approaching the Corpus Refinery.
• 2:58:45
  And the Technologies, we're trying there.
• 2:58:47
  And the reason we're making them aware of it is because we think
• 2:58:50
  they're fundamentally more scalable and as we proved them
• 2:58:52
  out. We plan to share that with them
• 2:58:55
  because As elon's. Like it's not really like we
• 2:58:57
  want to do these things. We're doing that because it's
• 2:58:59
  not happening fast enough. So if we can prove that it can
• 2:59:02
  be done faster, the intention is to transfer that knowledge to
• 2:59:05
  our large our current suppliers and the same is actually true,
• 2:59:09
  it was actually a clay process that we were playing with and we
• 2:59:12
  continue to work on. Same is true on that.
• 2:59:15
  So we've worked with our suppliers as well on trials and
• 2:59:18
  and we're sharing knowledge there.
• 2:59:20
  And the intention is just a help, the whole world, do this
• 2:59:23
  better. And ultimately the this here is
• 2:59:26
  is getting the lithium out or whatever it is out of the or And
• 2:59:32
  we're obviously building a cathode processing facility.
• 2:59:37
  Just adjacent to this building. So little further down the road,
• 2:59:41
  you'll see another large construction that's for cathode
• 2:59:44
  refining. I would like sit really prefer
• 2:59:50
  others to that. We're doing it because we have
• 2:59:53
  to not because we want to. Yeah, and in that case, there
• 2:59:55
  just isn't really any large-scale cathode production
• 2:59:58
  in the United States, and it needed to be done.
• 3:00:01
  And again, if we're going to do it, we're going to try to do it
• 3:00:03
  from first principles perspective.
• 3:00:04
  So we're we have tried a bunch of new things.
• 3:00:06
  There were confident that they will work, and as they prove out
• 3:00:10
  again, we want to bring them back to our suppliers that they
• 3:00:12
  can build new facilities more quickly with lesson.
• 3:00:16
  Investment. All right.
• 3:00:20
  Let's go to bed. Hi.
• 3:00:21
  Ben. Caleb are so similar on the
• 3:00:25
  renewable side. Is there more that you can do in
• 3:00:29
  Tesla to nudge the rest of the renewable industry to speed up
• 3:00:33
  since it's such big? Pillar of the master plan 3?
• 3:00:40
  Well, I mean, I don't know what more we can do, but I can say
• 3:00:45
  that if this like entrepreneurs out there that wanted to have,
• 3:00:49
  like, a guaranteed chance of success, it would be refining
• 3:00:53
  lithium or anodes, and cathodes, or any materials whatsoever for
• 3:01:01
  lithium ion cells as no-brainer. Yep.
• 3:01:07
  Like I were doing everything we can.
• 3:01:11
  So yeah on the other renewable energy thing.
• 3:01:15
  The thing that we can do Tesla is the more we reduce the cost
• 3:01:19
  of storage there, more reduce the cost of stationary storage
• 3:01:22
  and the more we bring like flexible load to the Grid in the
• 3:01:26
  form of like cars charging at the right times, the more
• 3:01:28
  valuable Renewables are because in like Texas right now.
• 3:01:32
  I wasn't joking like there's a lot of wind and at Night, wind
• 3:01:36
  is there's like almost too much wind because they don't want to
• 3:01:39
  turn the nukes off or whatever. So, so so so, so bringing really
• 3:01:45
  Low Cost Storage onto the grid makes Renewables more valuable,
• 3:01:48
  which ultimately will accelerate their deployment.
• 3:01:51
  So that's how we focus on it, I guess.
• 3:01:55
  Yeah, felipa. Thank you.
• 3:02:06
  It's CD, preferred Jeffries. I've got two questions.
• 3:02:10
  The first one. When I think about in this
• 3:02:13
  industry, everybody wants to be Tesla.
• 3:02:15
  Every car maker is trying to emulate what you're doing.
• 3:02:18
  Well, done. The one thing they don't do is
• 3:02:23
  focus on having as much growth as possible with as humans as
• 3:02:27
  possible. So I'm just trying to understand
• 3:02:29
  as you aim for 20 million units in 2030.
• 3:02:33
  How many models do you think? You need to get there and how
• 3:02:37
  does it fit into your drive to hyperscale?
• 3:02:41
  How do you manage this? And also the fact that probably
• 3:02:44
  consumers at some point, don't want to see a test.
• 3:02:47
  Allow the same test that every street corner.
• 3:02:49
  So I'm just trying to get your your sense of that.
• 3:02:52
  And my other question is on, on bi-directional.
• 3:02:56
  I mean, you you talked a lot about making no world more
• 3:03:00
  Renewables and better usage of cars.
• 3:03:03
  I think bi-directional charging is one way of better.
• 3:03:05
  Using cars, but you seem to be reluctant about doing that in
• 3:03:09
  the past. So I'm just wondering with or
• 3:03:11
  latest views are on the topic. Sure, I'm bi-directional it.
• 3:03:23
  It wasn't like a conscious decision to not do it.
• 3:03:26
  It just wasn't a priority at the time I think is maybe the way to
• 3:03:28
  think about it as we looked at as we continue to improve the
• 3:03:32
  Power Electronics. In our vehicle.
• 3:03:34
  We've found ways to bring bra directionality while actually
• 3:03:37
  reducing cost of Power Electronics in the vehicle.
• 3:03:40
  And as at all things Tesla, the the goal is usually to get more
• 3:03:45
  for less. And so we are in the middle of
• 3:03:48
  kind of like a Power Electronics.
• 3:03:50
  Retool, I would say that will bring that functionality. 82
• 3:03:52
  arvo all of our vehicles over the next you know two years
• 3:03:56
  let's say but but it's yeah I guess that's how I'll say there.
• 3:04:02
  Yeah I don't think very many people are going to use
• 3:04:04
  bi-directional charging unless you have a power wall because if
• 3:04:07
  you unplug your car your house goes dark and this is it be
• 3:04:11
  extremely inconvenient? Yeah, most of the value it comes
• 3:04:17
  in charging the car at the right time.
• 3:04:20
  It's not really about sending energy the other way.
• 3:04:23
  Yeah, I mean if you have a power wall that can take the house
• 3:04:27
  load then you can use the your car as Supple as a supplementary
• 3:04:31
  energy source to power wall and then you know, you're not going
• 3:04:35
  to drive everyone Crazy by unplugging your car and having
• 3:04:39
  the house go off. So I think there's some value
• 3:04:42
  there are as a supplemental energy source down the road.
• 3:04:47
  where if you have a powerwall you've not diminished the
• 3:04:52
  Convenience of the people in the house.
• 3:04:57
  And the question of number of models.
• 3:05:01
  Sorry what oh how many models of vehicles, not that many really
• 3:05:07
  tan? I don't know what that money.
• 3:05:11
  There's I mean what's happened with conventional cars is people
• 3:05:15
  would run out of things to do. So you want to thanks to do,
• 3:05:18
  they just end up reshuffling the deck and you have pretty much
• 3:05:20
  the same. I mean, how many variants of car
• 3:05:24
  are there on the road? There's like hundreds.
• 3:05:29
  But there are they good variants?
• 3:05:32
  No I mostly not there just variance for the sake of
• 3:05:35
  variance and but look at how is have things converged with the
• 3:05:40
  phone and when they used to be hundreds of flip phones, now
• 3:05:44
  what do we have? It'll be like that.
• 3:05:51
  George George generic is from canaccord genuity just had a
• 3:05:56
  question about your plans to grow market share in China.
• 3:06:01
  And also whether or not the political tensions between
• 3:06:04
  United States and China impact your long-term Ambitions their
• 3:06:08
  thank you. Hey, Tom don't don't sweat too
• 3:06:17
  hard. Well, they're still growing our
• 3:06:25
  market share in China, quite a strong actually.
• 3:06:29
  Especially, you know, earlier this year, we had a price
• 3:06:33
  adjustment after that, we actually generated a huge demand
• 3:06:37
  more than we can produce really. And as be long said, as long as
• 3:06:43
  your offer, a product with value at affordable price, you know,
• 3:06:52
  you don't have to worry about demand.
• 3:06:53
  That's basically the flow velocity that we, we follow, we
• 3:06:57
  try everything to cut cost from supply chain, from improve the
• 3:07:02
  efficiency in the factory, and the passed down that value to
• 3:07:05
  our customers. I think you guys that was, as
• 3:07:08
  long as we continue to do that, I'm not too concerned about The
• 3:07:10
  market share in China. And the second part
• 3:07:13
  geopolitical. No one else we do our best.
• 3:07:17
  We actually create a lot of jobs to the local community, and we
• 3:07:22
  generated also with, at our suppliers factories or will
• 3:07:29
  create a lot of jobs as well and will contribute a lot of to the
• 3:07:34
  local economy. I think you know as long as
• 3:07:37
  we're needed in this country I don't and see there much of the
• 3:07:41
  risk of that. Colin Colin Rush from
• 3:07:46
  Oppenheimer, you know, one of the things I was struck by in
• 3:07:49
  the presentation was the the operational efficiency metrics.
• 3:07:52
  You guys are talking about, could you talk a little bit
• 3:07:54
  about the velocity of learning cycles and how you guys track
• 3:07:57
  that? And think about that as an
• 3:07:59
  organization as you work into a variety of other areas from from
• 3:08:02
  an innovation perspective. Anyone want to try that?
• 3:08:11
  I mean I'll say one very high level thing which is you can't
• 3:08:14
  improve something you don't measure and we're like ruthless
• 3:08:17
  measures at Tesla and once you start measuring things that
• 3:08:23
  contribute to operational efficiency, you actually have a
• 3:08:26
  path to doing something about it.
• 3:08:28
  So I guess the key is a good measuring stick.
• 3:08:34
  Okay, but it's something I should say with respect to
• 3:08:40
  demand which I've said a few times over the years but is
• 3:08:43
  sometimes it needs to be said again which is the
• 3:08:51
  overwhelmingly. The there's the the desire for
• 3:08:56
  people to own a Tesla is extremely high.
• 3:08:58
  The limiting factor is their ability to pay for a Tesla?
• 3:09:02
  Not do they want? A Tesla, it's it's easy for
• 3:09:07
  people in this room to lose sight of that, if your income is
• 3:09:11
  far in excess of what a car costs, then you look at value
• 3:09:14
  for money, but you do not consider affordability.
• 3:09:17
  But for the vast majority of people, it is affordability
• 3:09:21
  driven. This is why we cannot simply
• 3:09:23
  double the price of the car or you could say and think about
• 3:09:27
  things in the limit where if you hadn't, you know, infinitely
• 3:09:31
  desirable car. But it costs Ten million
• 3:09:34
  dollars, it wouldn't matter because he was not kind of
• 3:09:38
  possible, 30 not have that. So demand is very much, a
• 3:09:44
  function of affordability, not desire, very important.
• 3:09:49
  One of the things we weren't sure about was the price
• 3:09:52
  elasticity of demand for Tesla. So like as we lower the price
• 3:09:58
  how much this demand increase, and we found that Even small
• 3:10:05
  changes in the price have a big effect on demand, very big.
• 3:10:09
  So that was a good thing to learn.
• 3:10:14
  Yeah. And then as of the sort, Hannah
• 3:10:18
  me question is very, very big because you could potentially
• 3:10:24
  have. I don't know, five times the
• 3:10:26
  utility of the asset that you currently have.
• 3:10:29
  So of passenger car is 10 or 12 hours a week of usage.
• 3:10:35
  Plus a lot of parking expenses, an autonomous car could be 50.
• 3:10:41
  Just 60 hours a week or something like that and you
• 3:10:46
  could get rid of a lot of parking expenses.
• 3:10:49
  So You know, if this if this is true, then as autonomy is
• 3:10:57
  effectively turned on for the fleet, it may be, the probably
• 3:11:02
  will be the biggest asset value increase in history overnight.
• 3:11:09
  Emmanuel. Thank you so much.
• 3:11:16
  I mean you're awesome from Deutsche Bank so as you start
• 3:11:20
  launching this next Generation vehicle and ramping up volume,
• 3:11:24
  what will be your nearest term priority in terms of segments or
• 3:11:29
  Vehicles Focus? The the slide that you showed
• 3:11:32
  with two vehicles on the rap seemed based on form factor.
• 3:11:35
  One of them maybe looks like a van and other one looks like
• 3:11:37
  maybe like a smaller vehicle like potentially a model to I
• 3:11:41
  guess. What is the nearest?
• 3:11:42
  Mm Focus for you in terms of ramping?
• 3:11:44
  Up the next Trend vehicle and how do you make sure that by
• 3:11:48
  lowering the price point? So much?
• 3:11:50
  Because the cost is going down, 50% you're not cannibalizing.
• 3:11:54
  The man and you know for your existing vehicles Demand for our
• 3:12:05
  vehicles. In terms of desire to own them
• 3:12:07
  may as well be infinite through. It's indistinguishable from
• 3:12:12
  infinite at this point. Forward ability is what matters.
• 3:12:16
  So as you make the car more affordable, we will have to
• 3:12:20
  meant go crazy. Basically, the issue is, how do
• 3:12:24
  we build the cars? The hard part is building the
• 3:12:26
  cars, I can't emphasize that enough.
• 3:12:28
  The hard part is building the cars and the entire supply chain
• 3:12:31
  that goes with the cars. This is a Logistics challenge of
• 3:12:38
  extraordinary difficulty. All the things that have to go
• 3:12:44
  into the car have to scale with the car, while everything is
• 3:12:47
  doing in an exponential ramp. And if you must even one of
• 3:12:52
  those things, Doesn't matter why?
• 3:12:56
  Earthquake flood fire Revolution.
• 3:12:59
  I thought I've heard them all. I mean, is any part of that
• 3:13:04
  supply chain gets interrupted your now healing, you have a
• 3:13:07
  seizure. The hard part is pulling the
• 3:13:14
  cars by far. and the supply chain that goes with it, I make
• 3:13:20
  you guys went to about Supply chains, do They might, you know,
• 3:13:26
  and the presentation had mentioned Perfection as a
• 3:13:29
  passing grade. We really need everything to
• 3:13:32
  happen perfectly and the strategy for mitigating the
• 3:13:35
  different risks. Some of which were anticipated.
• 3:13:37
  And some weren't is really bespoke to the situation.
• 3:13:40
  And and then really the you need subject matter expertise and a
• 3:13:45
  deep knowledge of the particular supply chain.
• 3:13:47
  You're managing to come up with our strategies.
• 3:13:50
  So we've seen everything from, as Elon mentioned, you know,
• 3:13:53
  first the the tariffs that flew back and forth.
• 3:13:55
  Then the ocean Logistics issue a chip shortage covid floods.
• 3:14:00
  There was a fire and a Fab in Japan that knocked down.
• 3:14:04
  There was a massive coach spike in Malaysia for a lot of the
• 3:14:09
  chips, or the back end of a lot of chips as was down there and
• 3:14:12
  more. This was one that Elon was
• 3:14:15
  involved with as well. And yeah, that's it's
• 3:14:19
  nerve-racking but somehow it works Yeah, and all those
• 3:14:25
  difficult, I think we have been laying the foundation to be as
• 3:14:30
  intimate to are all the tears of the supply chain building that
• 3:14:34
  control and, you know, directing the different tiers of supply
• 3:14:38
  chain. That's the way we are trying to
• 3:14:41
  mitigate the risks that come with it.
• 3:14:44
  You know, again, dual sourcing triple sourcing having
• 3:14:48
  redundancy is the way we've been trying to mitigate for it.
• 3:14:54
  Yeah, and that's fine. When we think about Vehicles
• 3:14:57
  when you think about three wires and architecture SX.
• 3:15:00
  As an architecture, our next Generation platform is more than
• 3:15:02
  one segment and really were thinking about all the segments
• 3:15:05
  that are available, that we haven't captured and where the
• 3:15:07
  market would be, and designing it with our supply chain
• 3:15:10
  Partners, so that we can go quickly through those segments
• 3:15:12
  for where we need. But to elon's point, if you make
• 3:15:15
  a car desirable and affordable, you know, oftentimes it doesn't
• 3:15:21
  necessarily matter what segment it's in.
• 3:15:24
  Cuz it's one that you want and we've seen that with model 3,
• 3:15:27
  when a lot of people thought the sedan was not going to be a
• 3:15:29
  great hit but we sell tons of them.
• 3:15:31
  So the Next Generation platform is not one vehicle that is, it
• 3:15:35
  has multiple and it's on the segment that we will really try
• 3:15:39
  and focus on that affordability and desirability point more over
• 3:15:42
  than where we started. Yeah, I mean it's like I think
• 3:15:48
  it's an old saying like battles are won with tactics or is the
• 3:15:50
  one with Logistics Logistics challenges here are enormous.
• 3:15:55
  And when you start like being a very significant percentage of
• 3:16:00
  an industry, you can't / capacities, it's not realistic.
• 3:16:05
  And some of these things like you said like well do all sorts
• 3:16:09
  of trouble Source. You can do that maybe for small
• 3:16:11
  things but you can't do it for big things because you If
• 3:16:15
  you're, if you're trouble sourced and one of it's like
• 3:16:19
  having a plane with three engines where if any of the
• 3:16:22
  three engines fails you crash. It's like, you know, so you have
• 3:16:29
  to either over capacities, which drives your capex up and it has
• 3:16:33
  idle suppliers somewhere and big warehouses or you designed to
• 3:16:38
  some overage that's, you know, reasonable and then You have
• 3:16:45
  expedite costs because inevitably there's something
• 3:16:48
  goes wrong somewhere and you got to fly things around.
• 3:16:52
  So it's really just the rate of progress is the rate at which we
• 3:16:55
  are able to scale. A 10,000 Logistics problems,
• 3:17:01
  most significant of those is the cell production.
• 3:17:07
  And so, we actually deliberately try to overdo cell production or
• 3:17:13
  self Supply to have that exceed. What is needed in vehicles,
• 3:17:20
  because if it goes below what's needed in vehicles then, but
• 3:17:23
  with the factories stole. So, but then, what do you do
• 3:17:28
  with all these extra cells? It's like, well, okay, so the
• 3:17:31
  much easier thing to scale up and down, is it powerwall and
• 3:17:35
  mega pack out? Put stationary storage.
• 3:17:38
  So we can then over capacities in cells and packs and scale
• 3:17:45
  production of stationary storage, which is much easier to
• 3:17:49
  scale than vehicle production. So that's strategically, I think
• 3:17:54
  a good thing. Yeah, the capex for mega pack is
• 3:17:59
  Tiny compared to capex for vehicles and also mega pack
• 3:18:04
  demand is quasi infinite it. Basically as long as we are
• 3:18:08
  competitive with utilities we can tell as many mega like it.
• 3:18:14
  Yeah kwazii infinite demand for that.
• 3:18:17
  Really work multi-turret, many terawatt hours and we got a long
• 3:18:22
  way to go to get too many terawatt hours per year.
• 3:18:28
  Then, let's go to down. Thank you Dan Levy, Barclays.
• 3:18:35
  I think we know competitive Dynamics, differ, significantly
• 3:18:40
  by region, you know, cost Dynamics, differ by region.
• 3:18:43
  We saw that with you're trying to gigafactory, far superior
• 3:18:46
  cost of what you had a free month.
• 3:18:47
  I realize it was a new Factory but they're clearly different
• 3:18:50
  Dynamics. So to what extent are the the
• 3:18:53
  cost strategies that you've laid out today.
• 3:18:55
  Do they differ by region or or is there more of a global one
• 3:18:59
  size fits all approach on reducing cost?
• 3:19:10
  Yeah, I think we're pretty consistent on to the strategy
• 3:19:14
  here, we'll try to have as much localized as possible.
• 3:19:18
  Like in China 95 percent over 95% of our supply chain are
• 3:19:22
  localized all the way from a first-tier the second, theorem
• 3:19:25
  that the desert here that give us the generator, tremendous
• 3:19:31
  savings on Cox and Also, have a very localized early, labor
• 3:19:39
  force, and we have access to skilled labor force, in the
• 3:19:43
  region and younger daughter region in Shanghai, and
• 3:19:46
  particular referring to in and among over 30,000 of employees.
• 3:19:53
  In China, we have probably less than 20 experts.
• 3:19:58
  So, you know, very deep localization who have done
• 3:20:01
  their, to be able to Have such cost structure in.
• 3:20:06
  We try to replicate this approach elsewhere and in
• 3:20:11
  different glue factories. Of course, you know, A supplier
• 3:20:14
  base that are different. The you know labor market are
• 3:20:18
  different region by region the country by country.
• 3:20:21
  But we try our best to to localize.
• 3:20:25
  Yeah and I think that's that's absolutely.
• 3:20:30
  Give us an advantage to compete with the Auto EMS around the
• 3:20:35
  globe. Another thing is your, we have
• 3:20:39
  this direct selling mode. Like Zach said, you know, it was
• 3:20:42
  save a tremendous money, no packs as well.
• 3:20:45
  We have full control over all expenditures across the the
• 3:20:50
  company. And, you know, we don't spend
• 3:20:53
  money on marketing or advertising.
• 3:20:57
  You everything we saved will become, you know, the value.
• 3:21:03
  We can offer to the Consumers. So that part on that part, we
• 3:21:07
  also followed pretty consistent strategy over the years.
• 3:21:13
  I mean, certainly the manufacturing stuff that Colin
• 3:21:15
  Pete and myself talked about whether you make it here or you
• 3:21:18
  make it in Europe, or you make it an Asia, it applies
• 3:21:20
  everywhere. We think about that from the
• 3:21:22
  ground up, not not specific to any region.
• 3:21:27
  Yeah, just one thing I would add to Tom's point about
• 3:21:30
  localization and to Lars is point here but I would actually
• 3:21:33
  say that we're moving in a direction of more
• 3:21:35
  standardization in terms of our factories, and our processes,
• 3:21:38
  and our cost reduction approaches.
• 3:21:40
  So as we've been thinking about the Next Generation platform,
• 3:21:43
  we've been thinking about the volume that we aspire to build
• 3:21:46
  against that. How many individual Factories
• 3:21:49
  do? We need to build?
• 3:21:51
  And what is the fastest possible way to expand that footprint
• 3:21:55
  around the world? What we've done with model Y is
• 3:21:58
  each Factory is an incremental Improvement of the previous
• 3:22:00
  Factory which requires engineering hours engineering.
• 3:22:03
  Spend for each Factory design and then you end up with
• 3:22:06
  factories that are slightly different from each other and as
• 3:22:09
  we move towards the Next Generation platform I think the
• 3:22:13
  term you use Lars is copy paste. So to get it right from the
• 3:22:17
  first time, you know, certainly there will be things that we
• 3:22:21
  learn and we replicate and we make adjustments to, but try to
• 3:22:24
  have as much standardization in compensation as Possible, which
• 3:22:27
  allows us to go as quickly as possible with expansion.
• 3:22:33
  Oh, Veronica on the side. Thanks, two questions.
• 3:22:40
  One for exact one for a shock self.
• 3:22:42
  Exact. How do you think about a
• 3:22:44
  long-term growth and operating margins for the business?
• 3:22:49
  And when we look at 150 to 175 billion in capex as the
• 3:22:54
  remaining balance of that, is that essentially the capex
• 3:22:57
  guidance through 2030 for a shock, the multi trip
• 3:23:02
  reconstruction that you highlighted, how is that
• 3:23:04
  different from what mobilized does, what the REM map that they
• 3:23:08
  have? And can you give us any color on
• 3:23:10
  the AP for Hardware? So with respect to operating
• 3:23:18
  margins, you know we intend to continue to improve operating
• 3:23:22
  expenses as a percentage of Revenue over time and we're
• 3:23:25
  continuing to take cost out of our products and try to keep
• 3:23:29
  gross margins in a place that's healthy as well.
• 3:23:32
  And so, you know, from the hardware perspective of the
• 3:23:35
  business units are expectation, that we'll continue to stay in a
• 3:23:38
  healthy place over time. And then there's a software
• 3:23:42
  portion that's added on top of that.
• 3:23:44
  And so Ilan has commented on this many times about the impact
• 3:23:47
  that full self-driving software can have on the economics of the
• 3:23:50
  business. And in that space, you know,
• 3:23:53
  profitability operating margins would be impacted dramatically
• 3:23:57
  specifically to your point about capex guidance.
• 3:24:00
  You know, the intent of that slide was not to provide
• 3:24:02
  specific guidance, but to just be transparent about internally,
• 3:24:06
  what are rough estimates are and to provide context that We think
• 3:24:11
  getting to 20 million vehicles per year, and one terawatt hour
• 3:24:14
  of energy storage is very feasible, relative to our
• 3:24:17
  expected cash generation for the business.
• 3:24:19
  That's, that's a number that will move.
• 3:24:21
  As we learn, we may choose to vertically integrate more into
• 3:24:24
  things. We may find efficiencies
• 3:24:26
  elsewhere to the whole conversation.
• 3:24:28
  We were having earlier about, do we do mining?
• 3:24:30
  Do we not do mining, but the entire point is to say that this
• 3:24:34
  is something that's entirely possible based upon our
• 3:24:36
  forecast, All regarding the multitude pre-construction.
• 3:24:41
  My point I was trying to make was that we Wanda Auto label,
• 3:24:46
  most of the data and we can collect data from the fleet and
• 3:24:48
  then reconstruct local areas that can act as supervision for
• 3:24:51
  those clips since we want to build a scalable self-driving
• 3:24:55
  system, we don't want to rely on HD Maps or anything.
• 3:24:57
  Even though we could take trivially build something using
• 3:24:59
  the same technology, but this is mostly for auto labeling.
• 3:25:04
  And this provides like precise 3D labels for all the video
• 3:25:07
  sequences involved in the Suction.
• 3:25:13
  It's something that I don't know if we touched on this much.
• 3:25:17
  But in terms of training, we have one of the biggest
• 3:25:21
  neural-net Training Systems in the world and that we expect to
• 3:25:26
  increase that capability by an order of magnitude by the end of
• 3:25:28
  this year, and probably another order of magnitude by the end of
• 3:25:34
  next year. With some combination of Nvidia
• 3:25:38
  and Dojo. So I'm not sure the scale of
• 3:25:43
  that is quite going down, people, its enormous.
• 3:25:50
  Maybe the final two questions, one from Alex and one from
• 3:25:54
  Chris. Okay, Alex Potter with Piper so
• 3:26:02
  I definitely want to ask Drew or anybody else up there for an
• 3:26:07
  update on dry battery electrode, right?
• 3:26:09
  So, if you're try to over capacity as towards cells scale,
• 3:26:14
  a lot of this, that clearly there's a lot of moving pieces,
• 3:26:17
  and it's a complex sort of Orchestra of, with the supply
• 3:26:19
  chain. But a lot of it comes down to
• 3:26:21
  drive battery electrode, at least to me.
• 3:26:24
  So, how are you trending? That was the most fascinating
• 3:26:27
  part of the factory tour for me and looking through that window.
• 3:26:29
  Thing that this is clearly not a science project but anything
• 3:26:33
  that you're willing to disclose yields progress where you are
• 3:26:37
  today versus where you thought, you were going to be.
• 3:26:40
  Yeah, thanks. Yeah.
• 3:26:49
  I mean as you saw right like it's this is this is a real
• 3:26:51
  factory making a lot of dry electrode in an automated
• 3:26:55
  fashion. We've made a lot of progress.
• 3:27:00
  It's it's a, it's a spectrum like, we're perfectionists and
• 3:27:05
  we have clear and goals and we are every week that goes by
• 3:27:11
  making progress towards those Zen goals, whether it's speed of
• 3:27:14
  the tool yield of the of that process or the downstream
• 3:27:17
  process we have installed out yet on the rate of progress
• 3:27:24
  either and that's both on the anode and the cathode side.
• 3:27:28
  And I think the great thing about where we are with the
• 3:27:32
  overcapacity that Elon mentioned is it's giving us the
• 3:27:35
  opportunity to experiment as we go rather than just being like
• 3:27:40
  stuck to something that we happen to.
• 3:27:43
  Off a year and a half ago and it is here something that you do on
• 3:27:46
  has said to the team. Many times is it okay to scrap
• 3:27:50
  equipment or money? It's not okay, to scrap time.
• 3:27:54
  So the way we've been approaching it as
• 3:27:56
  probabilistically. What do you think?
• 3:27:58
  Is the most likely thing to succeed and even actually in the
• 3:28:02
  factory? Here you saw more than one, an
• 3:28:04
  old line that they are actually operating two slightly different
• 3:28:07
  versions of the the final process step of the powder
• 3:28:10
  entering the tool, and it's a competition.
• 3:28:13
  Don't you know which is going to be more higher yield which is
• 3:28:15
  going to perform better and we have the luxury to be able to do
• 3:28:17
  that. And we're taking full advantage
• 3:28:20
  of it to advance the technology as quickly as possible.
• 3:28:24
  The dry electrode problem is is really quite a hard problem.
• 3:28:28
  I mean, we acquired Maxwell really just for the dry
• 3:28:30
  electrode technology, but just just illustrates what a gigantic
• 3:28:36
  Gap there is between something working at small scale and at
• 3:28:39
  Large Scale. We've had a very extremely
• 3:28:44
  talented team of Engineers working on scaling, the dry
• 3:28:49
  electoral process and having to be reliable and consistent.
• 3:28:54
  And we've been grinding hard literally and figuratively on
• 3:28:58
  this for quite a while. It seems likely that we will be
• 3:29:04
  able to scale it to volume this year.
• 3:29:07
  Yeah. Yeah.
• 3:29:08
  I mean we're basically increasing the output week over
• 3:29:10
  week, like roughly 1 K week per quarter is our internal Target
• 3:29:16
  and we're you know, we're tracking two that I think the
• 3:29:20
  what you saw is effectively like a ton of material per hour per
• 3:29:25
  tool, it's kind of like hard to like Rationalize, like what that
• 3:29:30
  really means. But it's it's it is, it's
• 3:29:35
  different than like, oh something works in a lab.
• 3:29:37
  It's like, well, when it's tons of material per hour, there's
• 3:29:40
  just different kinds of problems.
• 3:29:41
  Like, even if you have like, point one percent Escape of
• 3:29:45
  like, finds into the enclosure that your equipment is in.
• 3:29:50
  Now, you have a dust problem and you and like that could short
• 3:29:53
  out some Electronics. It's like just things like this,
• 3:29:56
  where when you're on a lab scale, do not even notice it.
• 3:29:58
  But When you're doing thousands of tons over the course of
• 3:30:01
  months, it's like, oh, a new failure mode, we found and
• 3:30:03
  that's where we're at, but we're not going those out though and
• 3:30:08
  and the team is grinding through through it.
• 3:30:11
  But, but progress every week The last question from Chris.
• 3:30:18
  Thanks for taking the question. I have a few follow-ups on the
• 3:30:21
  next gen vehicle first. When do you think we'll get a
• 3:30:24
  look at it? Maybe a prototype second, are
• 3:30:28
  there any details that you think you can share in terms of the
• 3:30:31
  size, the content, the performance, and then third, I
• 3:30:34
  think you mentioned that you would produce it in other
• 3:30:37
  plants. In addition to Mexico, should we
• 3:30:39
  take that to mean that you can launch it at an existing plant
• 3:30:42
  before you're finished constructing, the new plant in
• 3:30:44
  Mexico? I think we'll actually have to
• 3:30:49
  probably decline that answer. We will have a proper sort of
• 3:30:56
  product event but that would be jumping the gun.
• 3:30:59
  If we're answer your questions, maybe another question.
• 3:31:02
  If this? Yeah.
• 3:31:05
  Yeah, I don't know anyone. well, well then of I have two
• 3:31:14
  questions. This has really been a very
• 3:31:15
  impressive afternoon. Thank you so much, Ilan.
• 3:31:18
  I'm curious as you've doubled and you'll double again, how
• 3:31:22
  what you've learned about sort of managing a larger Enterprise,
• 3:31:26
  and what you might have to, you know, do to manage a bigger
• 3:31:30
  Enterprise and then, secondly, I'm curious on your thoughts on
• 3:31:34
  how, you know, generative Ai. And, you know, these rapid
• 3:31:38
  breakthroughs in Ai and the last month's could help you.
• 3:31:42
  Make cars sort of you know less hard to make.
• 3:31:47
  Thank you. I don't see a I helping us make
• 3:31:53
  cars anytime soon at that point I mean no one any of us working.
• 3:32:01
  I think big problems, she'll just chill out.
• 3:32:07
  I mean, I'm a little worried about the AI stuff.
• 3:32:14
  I think it's something, I don't know, which we should be
• 3:32:18
  concerned about. I don't think we should meet
• 3:32:24
  some kind of like regulatory Authority or something.
• 3:32:27
  It's over seeing AI development and just making sure that its
• 3:32:31
  operating within the public interest.
• 3:32:32
  And, you know, it's quite a dangerous, quite dangerous
• 3:32:37
  technology, and I fear, I may have done some things to
• 3:32:42
  accelerate it, which is, I don't know.
• 3:32:47
  So, I mean some of the AI stuff I think it's just obviously
• 3:32:53
  useful, like what we're doing with self-driving, which is, you
• 3:32:59
  know, some people think is an AGI type problem.
• 3:33:03
  I don't think it's quite an AGI problem, but it's certainly
• 3:33:06
  requires very sophisticated neural net because the road
• 3:33:12
  system is designed for eyes and biological neural Nets.
• 3:33:16
  So naturally the analog to that is cameras with digital during
• 3:33:24
  Nets and I think we found is just, there's a see, if you
• 3:33:33
  actually look at our neural net architecture in the car, it's
• 3:33:36
  kind of insane. Frankly, that's the pain that's
• 3:33:40
  upon. That's upon Nets.
• 3:33:42
  The visualization told crashes in the open.
• 3:33:45
  Yeah, it's so complicated. That no one can visualize this
• 3:33:47
  right now. Yeah, it's hard to visualize.
• 3:33:49
  Just like literally, it looks insane.
• 3:33:52
  So now I guess something like that's happening in our brains
• 3:33:55
  while we drive around, which is pretty well.
• 3:33:59
  So, you know, I don't know. Like Tesla's doing good things
• 3:34:09
  and AI. This one stresses me out.
• 3:34:16
  So I don't know what to think she say about it.
• 3:34:20
  You know. Okay, given it 7:00.
• 3:34:25
  I think that's all the time. We have.
• 3:34:27
  Thank you very much for coming and thank you.