Latest / Mechanical Engineering Made Simple / (#21) Early Decisions, Lifetime Costs: How Lean Engineering Masters Risk.

Transcript

(#21) Early Decisions, Lifetime Costs: How Lean Engineering Masters Risk. · Mechanical Engineering Made Simple

• 0:00
  Welcome back, design junkies, to Mechanical Engineering Made
• 0:02
  Simple. Today we are launching a deep
• 0:04
  dive into lean engineering principles, but we're looking
• 0:08
  way beyond the usual factory floor stuff.
• 0:10
  We're really digging into how Lean applies to product
• 0:13
  development. You know, the whole enterprise
• 0:15
  structure. That's right.
• 0:16
  And for you, the early stage engineer listening, the core
• 0:19
  idea of Lean, getting rid of waste, really focusing hard on
• 0:24
  value creation and well, speeding U cycle time.
• 0:28
  Fundamentally, it's about enabling better engineering,
• 0:30
  less risky engineering. Our mission today really is to
• 0:34
  show you why those decisions you make right now, like in the
• 0:36
  conceptual design phase, they're actually creating the biggest
• 0:39
  risks and locking in the biggest costs way down the line.
• 0:43
  Yeah, And we're going to unpack some concepts that might feel a
• 0:45
  bit backwards at first, like how leveling out production volumes
• 0:48
  for your suppliers using something called Hi Jenka, is
• 0:51
  actually one of your most powerful tools.
• 0:52
  Even if your day job is all digital design, right?
• 0:56
  The big picture is Lean aims to create value for the customer.
• 1:00
  It's not just about slashing internal costs.
• 1:02
  OK, so let's first just define our playground here.
• 1:07
  What exactly is product development PD?
• 1:10
  Our sources define it pretty broadly.
• 1:12
  It's basically the whole set of activities, starting from when
• 1:16
  someone sees a market opportunity all the way through
• 1:19
  until the product is actually made, sold and delivered
• 1:21
  successfully. When you put it like that, it
• 1:24
  sounds less like, you know, creative bursts and more like a
• 1:27
  huge information management job. Precisely, yeah.
• 1:30
  If you think about it over time, product development is
• 1:33
  fundamentally about gathering information efficiently.
• 1:35
  And this is key, eliminating risk.
• 1:38
  Think of four things moving over a project timeline.
• 1:41
  Time moves forward, information builds up, risk hopefully goes
• 1:44
  down, and eventually value realized goes up.
• 1:46
  PD is really disciplined to risk reduction.
• 1:49
  And risk and design isn't just one thing, is it?
• 1:51
  It comes in like different flavors of uncertainty.
• 1:53
  Absolutely. It's helpful to categorize this
• 1:55
  uncertainty. There's the simple lack of
• 1:58
  knowledge, stuff that just don't know yet.
• 1:59
  Then there's lack of definition, that sort of fuzzy target we
• 2:02
  often start with. And then the difference between
• 2:04
  your known unknowns, things you know you need to figure out, and
• 2:07
  the really nasty ones, the unknown unknowns, the surprises.
• 2:11
  OK, let's dig into this next bit because it feels super important
• 2:14
  for anyone in design, especially early on.
• 2:17
  Why do those decisions made when you're basically just searching
• 2:20
  ideas matter so much more than, say, the thousands of tiny
• 2:23
  decisions later in detailed design?
• 2:25
  Well, it comes down to this non linear trade off.
• 2:28
  It's a balance right between how easy it is to change the design
• 2:33
  versus how much of the total life cycle cost the LCC you've
• 2:36
  already sort of committed to in the very early stages,
• 2:40
  conceptual preliminary design. The ease of change is almost
• 2:42
  100%. Realize the material won't work,
• 2:44
  Swap it out relatively easy. But that life cycle cost
• 2:48
  commitment that shoots up really fast at the same time, doesn't
• 2:50
  it? Rockets up, Yeah, while the ease
• 2:53
  of change only dips a bit during preliminary design, that LCC
• 2:56
  committed cost curve just takes off.
• 2:59
  You haven't spent most of the money yet, not even close.
• 3:02
  But your early choices, the architecture they've already
• 3:05
  locked in most of the total cost the product will ever see,
• 3:08
  everything from materials to maintenance, even disposal.
• 3:11
  OK, that makes sense. Like if I pick some super exotic
• 3:14
  composite early on or a really tricky machine part, right?
• 3:17
  I've just committed the project to paying that premium for
• 3:20
  potentially decades. But wait, if LCC commitment
• 3:23
  jumped so fast, wasn't that tempt teams to just put off
• 3:25
  decisions, delay commitment as long as possible?
• 3:28
  Can that cause its own kind of waste?
• 3:30
  Yeah, that's exactly the tension Lean tries to manage.
• 3:33
  Delaying commitment is what traditional places often do, but
• 3:36
  that just leads to endless iteration waste and blows up the
• 3:39
  schedule. Lean pushes for disciplined,
• 3:42
  smart choices upfront because avoiding the commitment is often
• 3:45
  way more costly than committing wisely early on.
• 3:48
  Once you get into detail design, making changes gets dramatically
• 3:51
  harder. Ease of change might drop to
• 3:53
  80%, maybe even 66% later. Those early choices really
• 3:58
  hammer home while you need disciplined, lean thinking right
• 4:01
  from the start. OK.
• 4:02
  So we get the huge impact, the risk of those early decisions,
• 4:05
  but avoiding that risk takes more than just, you know, good
• 4:09
  intentions. It needs a system, a tool kit.
• 4:12
  We usually think of the Toyota Production System, TPS and maybe
• 4:14
  Kaizen Continuous improvement. Kaizen's great.
• 4:17
  It's important, but it's only one piece.
• 4:19
  TPS is really a whole toolbox for getting rid of waste or
• 4:21
  muda, and we can totally apply these principles to design work
• 4:25
  even if we never touch a factory floor.
• 4:26
  All right, let's breakdown some key tools from that kit, but
• 4:29
  frame them for a designer, not a factory manager.
• 4:32
  OK, let's start with stability. There's Hijinka.
• 4:35
  This is all about leveling out production volume and the mix of
• 4:38
  products. It basically stabilizes the
• 4:41
  rhythm of the whole operation, the supply chain too, and that
• 4:44
  minimizes design chaos caused by unpredictable demand signals
• 4:48
  bouncing around and. What about stopping mistakes
• 4:50
  before they even happen? That seems crucial for
• 4:52
  designers. Right, that brings us to the
• 4:55
  quality side. We have judoka Think of it as
• 4:58
  designing systems that automatically detect problems
• 5:01
  and stop built in error flags. If something goes wrong, the
• 5:04
  system stops itself and signals hey problem here and that pairs
• 5:09
  with Pokyoke. That's basically mistake
• 5:10
  proofing. Foolproof prevention for a
• 5:12
  designer Pokyoke means building constraints right into the
• 5:15
  product or process so mistakes are physically impossible.
• 5:18
  Like, you know, making a connector asymmetrical so it
• 5:21
  only plugs in the right way. OK, that's a very direct
• 5:24
  practical design example. What about flexibility?
• 5:26
  How does that fit in? That's Shojenka.
• 5:28
  It means flexible task assignment, being able to shift
• 5:31
  people and resources around quickly to deal with changes in
• 5:33
  volume or to improve productivity.
• 5:36
  For design team this could mean having cross trained engineers
• 5:39
  who can jump between different subsystems depending on where
• 5:41
  the bottleneck is right now. And finally, there's 5S
• 5:45
  cleanliness order discipline in an engineering context that's
• 5:49
  more about things like solid data management, clear version
• 5:52
  control, keeping the digital workspace tidy.
• 5:55
  So sticking with this idea of cutting muda in the design
• 5:58
  office, where does this waste reduction show up most
• 6:01
  powerfully in terms of like components and the overall
• 6:04
  product structure? This is where Lean really gives
• 6:06
  you a competitive edge. Design reuse or multi use parts.
• 6:10
  It's all about standardizing the underlying bits and pieces.
• 6:13
  Stop reinventing the wheel. Eliminate the waste of creating
• 6:16
  unique solutions when a perfectly good one already
• 6:19
  exists. Sounds like common sense almost,
• 6:21
  but the examples you mentioned show the impact can be enormous.
• 6:24
  Oh, it's huge. Think about commercial jets
• 6:26
  using the same main engine starter across different models,
• 6:29
  say A747400A767. Suddenly the airline needs way
• 6:34
  fewer spare parts to service those planes.
• 6:36
  Big savings. Simply logistics.
• 6:38
  Or another really powerful 1 military helicopters getting
• 6:42
  something like 85% parts commonality between the UH1Y and
• 6:46
  the H1Z models that radically cuts down the maintenance, crew
• 6:50
  size, the spare parts needed in the field.
• 6:52
  Massive operational advantage. Wow.
• 6:55
  Yeah, huge cost avoidance and gains just from standardizing
• 6:58
  parts early on. But if we're really serious
• 7:00
  about lean, we can't just stop at the parts list, right?
• 7:03
  We need to look at the whole organization creating the value.
• 7:05
  This takes us to the lean extended enterprise idea.
• 7:08
  Exactly. Lean thinking can't just live on
• 7:10
  the factory floor or even just within the product development
• 7:12
  team. It has to stretch out
• 7:13
  horizontally, you could say to include the whole lean, extended
• 7:17
  enterprise, suppliers, partners and yeah, absolutely critically
• 7:20
  the customers too. OK, but when the scope gets that
• 7:23
  wide, defining who actually matters, who the stakeholders
• 7:26
  are becomes really important. The lean definition needs to
• 7:30
  focus on value creation I guess. It does for the lean enterprise.
• 7:33
  A stakeholder is basically any group or individual who directly
• 7:37
  or indirectly effects or is affected by how well the
• 7:41
  enterprise creates value. See how that shifts the focus?
• 7:44
  It's not just about profit for shareholders, it's about
• 7:46
  optimizing the value streams themselves.
• 7:49
  In figuring out which stakeholders are most relevant,
• 7:51
  there's salience. You use three main attributes.
• 7:54
  These govern how you interact with them.
• 7:56
  That's right. Those three are power,
• 7:58
  legitimacy, and criticality. So power is just their ability
• 8:01
  to influence things, could be coercive, could be based on
• 8:04
  resources or even symbolic. Legitimacy is about whether
• 8:07
  their relationship or their claim is seen as appropriate, as
• 8:10
  valid. And criticality is about timing.
• 8:12
  Is their claim urgent? Is it super important for the
• 8:14
  enterprise to keep running? Managing all these complex
• 8:17
  relationships is well, it's central to making lean work
• 8:20
  across the board. That complexity sounds like
• 8:23
  exactly why collaboration is often so difficult.
• 8:26
  We have that great case study looking at engineering
• 8:29
  integration moving from IC to RTCE.
• 8:32
  Integrated Concurrent engineering established the idea
• 8:35
  of getting everyone together, but real time concurrent
• 8:38
  engineering RTCE really forced the doing it together part.
• 8:42
  Yeah, ICE, which started at JPL back in the mid 90s, was
• 8:46
  foundational, but moving to RTCE was seriously ambitious.
• 8:50
  It shifted the whole focus from just talking about the design in
• 8:53
  meetings to actually doing the design work together
• 8:55
  simultaneously. All the design data flowed
• 8:58
  through a central server, so every single designer had the
• 9:01
  latest current information instantly.
• 9:03
  And the technical results were pretty mind blowing, weren't
• 9:06
  they? That cycle time drop.
• 9:07
  Absolutely radical. Yeah.
• 9:08
  RTCE got them to a point where they could create complex new
• 9:11
  spacecraft designs in maybe 4 hours, which used to take them
• 9:14
  something like 4 weeks. They could explore thousands of
• 9:17
  design options in real time because that whole back and
• 9:19
  forth iteration loop was just gone.
• 9:21
  OK, that's the amazing technical success story.
• 9:24
  But where did the organizational friction come in?
• 9:28
  What stopped it from being fully adopted?
• 9:30
  The people problem always comes back to that, doesn't it?
• 9:33
  Functional managers, you know, the heads of different
• 9:35
  departments, they were still incentivized to optimize their
• 9:38
  own little piece of the puzzle, their subsystem, not the whole
• 9:42
  system. So for instance, when internal
• 9:45
  R&D money got tight, these managers just weren't willing to
• 9:48
  use their limited funds to support what they saw as some
• 9:51
  other managers pet project, even if it benefited the whole
• 9:54
  program. That lack of aligned goals, the
• 9:56
  lack of clear funding authority, it created huge resistance.
• 10:00
  Silos went back up. So the core lesson is pretty
• 10:03
  stark. Silent thinking just kills
• 10:06
  systemic efficiency, whether it's in the design office or on
• 10:08
  the shop floor. Which brings us neatly back full
• 10:11
  circle to that systemic thinking lean demands, especially for
• 10:14
  stabilizing the supply chain with Hijinka.
• 10:16
  Right, We need to really grasp the basic problem.
• 10:19
  In most traditional supply chains demand variability
• 10:23
  amplification. You might know it as the
• 10:25
  bullwhip effect. It happens because every single
• 10:28
  step in the chain is busy trying to optimize its own little
• 10:31
  world. Local optimization can.
• 10:33
  You give us a quick, concrete example of how that local focus
• 10:37
  messes things up upstream. Sure, Imagine a trucking
• 10:40
  planner. Their main job metric is
• 10:42
  probably minimizing trucking cost per mile, right?
• 10:45
  So they have a strong incentive to only send out full trucks.
• 10:48
  That means they batch orders together.
• 10:50
  They minimize the mix of products on the truck to reduce
• 10:52
  handling. Now that's great for trucking
• 10:54
  efficiency, locally optimal. But that decision sends this
• 10:57
  jerky, amplified demand signal upstream to manufacturing.
• 11:01
  Suddenly manufacturing sees these huge, infrequent orders
• 11:04
  instead of a smooth flow. Causes chaos, inventory,
• 11:07
  balloons, just massive waste. Right.
• 11:09
  So the variability forced on the downstream folks makes the
• 11:12
  upstream operations constantly scramble to catch up with these
• 11:15
  big lumpy demands. Synchronized planning goes out
• 11:17
  the window. And the solution is value
• 11:19
  delivery, Hijunka. Instead of letting everyone do
• 11:22
  their own local otimization dance, the enterprise pushes A
• 11:26
  globally leveled signal down through the whole supply chain.
• 11:30
  Replenishments get supplied with a steady volume and a steady
• 11:33
  mix. For example, maybe the goal is
• 11:35
  always 4 trucks a day. To maintain utilization, you
• 11:38
  send 4 trucks, adjusting the mix slightly and predictably, even
• 11:42
  if daily demand wiggles a bit. OK, so.
• 11:44
  That's stability. That consistent leveled signal
• 11:48
  let's the upstream departments finally breathe and stabilize
• 11:51
  their operations. Precisely that steady drum beat
• 11:53
  allows upstream operations, finishing materials, prep,
• 11:56
  whatever to globally optimize their work.
• 11:59
  It dramatically cuts the inventory needed everywhere
• 12:01
  because things are just much better synchronized.
• 12:04
  The sort of conceptual goal, the end game of applying Hybrisca
• 12:06
  across the supply chain is to maybe even enable a ship every
• 12:10
  part everyday approach. By smoothing out all that
• 12:13
  replenishment variability in both volume and mix, you
• 12:16
  stabilize the whole ecosystem. Wow.
• 12:18
  OK, this has been quite a journey we went from the huge
• 12:21
  impact of your very first design sketch and that LCC commitment
• 12:24
  trap all the way through the need for system wide
• 12:28
  collaboration like our TCe even with its challenges and finally
• 12:34
  landing on the the incredible power of High Junka to tame
• 12:37
  variability way beyond just the factory.
• 12:40
  Yeah, and. The main take away for you?
• 12:41
  The design mine you're listening is probably this.
• 12:44
  Really applying lean thinking across the whole enterprise
• 12:47
  forces a pretty fundamental shift in how you see
• 12:49
  competition. If you're in a traditional
• 12:51
  company, you tend to think about competing firm against firm.
• 12:54
  But lean enterprises, well, they have to achieve their advantage
• 12:57
  by shifting the whole game. It's not about your firm
• 12:59
  anymore. It's about the capability and
• 13:00
  efficiency of your entire production network.
• 13:03
  That network, including your suppliers, your partners, your
• 13:05
  customers, if that's your real source of lasting value until
• 13:08
  the next. Equation design junkies.
• 13:10
  TuneIn every morning at 11. AMA.
• 13:12
  Yeah.