Latest / eMotors: Electric Revolution / Cutting the TCO: Autonomous Electric Fleets Reshape Urban Transit
Transcript
- 0:09OK, so let's unpack this. We are, I mean, we're really
- 0:12standing on the edge of a huge shift in transportation.
- 0:15Maybe the biggest is the car itself.
- 0:16Exactly. We're talking about the rise of
- 0:18autonomous electric vehicles robotaxis.
- 0:21And the promise here is, you know, safer, more accessible,
- 0:24and way, way cheaper mobility for everyone.
- 0:27That's why today's deep dive is really for anyone who's curious
- 0:30about this stuff about autonomous mobility on demand or
- 0:34a mod. And we're not really tackling
- 0:36the tech question of if the cars will work.
- 0:38We know they're coming. The real question is financial
- 0:41and logistical. Right.
- 0:44How do these fleets get so efficient that they can actually
- 0:47replace, say, a city bus system or even you owning your own car?
- 0:51To figure that out, we've got sources covering everything from
- 0:54the the deep economics of these fleets to the really complex
- 0:58algorithms that run. Them.
- 0:59The brain, essentially. Their brain and their central
- 1:01nervous system. And we'll even look at the, you
- 1:04know, the very real cultural pushback they're getting in some
- 1:06cities. So our mission today is to
- 1:09breakdown what's called the total cost of ownership, the
- 1:11TCO. Because that's what helps us
- 1:13understand why robo taxis might not just be a little bit
- 1:15cheaper, but could completely shatter the way we pay for
- 1:19transport today. OK, so here's where it gets
- 1:21really interesting. Let's start with the economic
- 1:23numbers because frankly they feel a little unbelievable.
- 1:27We're drawing on a big total cost of ownership analysis.
- 1:30It was a Tuchem pre study and it looked at deploying these 30
- 1:34passenger autonomous electric vehicles or AE VS.
- 1:38And they modeled it for a city like Singapore with a target
- 1:40date of 2030. What's fascinating here is that
- 1:43you kind of have to ignore the sticker price.
- 1:46Yes, absolutely. The initial cost to buy one of
- 1:49these AE VS is definitely higher than a conventional bus or van.
- 1:54Even in 20-30. Even by 20-30, the study
- 1:56projects that the full autonomy hardware suite, you know, all
- 1:59the sensors and processors might still add an extra, say $9000 to
- 2:04the vehicle cost. But that's a one time cost that
- 2:07just gets spread out over years and years of operation.
- 2:09Exactly. It fades away.
- 2:10OK. But a 75% reduction in cost?
- 2:14That sounds almost unbelievable. We've been promised cheap public
- 2:17transport for decades, and it always needs heavy subsidies.
- 2:21So why should we believe these numbers now?
- 2:24Is it really just about getting rid of the driver?
- 2:26It is primarily about the driver, and you know that
- 2:30skepticism is healthy because these are disruptive numbers,
- 2:33right? The study showed that these AE
- 2:36VS can reduce the TCO per passenger kilometer by up to a
- 2:41massive 75% compared to a regular van, whether it's gas or
- 2:45electric. 75% And what about compared to say a normal city
- 2:49bus? That's maybe even more critical
- 2:51for city planners. The reduction was up to 60%
- 2:54compared to a conventional single or double deck bus. 60%
- 2:57cheaper per person per kilometer.
- 2:59That completely rewrites the financial playbook for urban
- 3:02transit. It does, and if you ask what the
- 3:04single biggest driver of that is, the answer is labor.
- 3:06It's just simple math. So how much does a driver
- 3:09actually cost? Well, when you look at any
- 3:12piloted vehicle fleet of taxi a bus driver, compensation makes
- 3:17up somewhere between 40 and 70% of the total cost of ownership.
- 3:2140 to 70, that is a staggering range.
- 3:24It is. So removing the human isn't just
- 3:26saving a salary. It's the benefits, the
- 3:29scheduling, the limits on driving hours, all of it.
- 3:32Precisely. We'd already seen simpler
- 3:34studies on commercial robo taxis showing like a 30% TCO reduction
- 3:40just from taking the driver out. But now we're layering 2 things
- 3:43together, autonomy and electrification.
- 3:45Yes, and that synergy is key. OK.
- 3:47So let's talk about that. The E in AEV.
- 3:49How does making an electric compound those savings?
- 3:52You get 2 really big distinct financial benefits.
- 3:55The first one is maintenance. Right, fewer moving parts.
- 3:58Dramatically fewer electric powertrains.
- 4:01The batteries and motors are just so much simpler than
- 4:03internal combustion engines. The estimates we've seen put EV
- 4:06and a EV maintenance costs at about 50% lower than for a
- 4:10diesel or gas vehicle. Half the maintenance costs right
- 4:12off the top. That's a huge recurring saving.
- 4:14What's the second piece? Energy efficiency.
- 4:16It's how the vehicle is actually driven.
- 4:18Humans are not always the smoothest drivers.
- 4:22Even the best of us, we're erratic, we brake hard, we
- 4:25accelerate too fast. Autonomous vehicles can drive
- 4:28with near perfect optimization. OK.
- 4:30And that saves energy. The studies show Avs can consume
- 4:33about 10% less energy just from smoother driving and better
- 4:37routing. So if you add it all up, you get
- 4:41rid of, say, half the total cost by removing labor.
- 4:44Then you cut maintenance costs in half.
- 4:46Then you gain 10% on energy efficiency.
- 4:49Suddenly that 75% TCO reduction starts to feel very grounded in
- 4:53math, not just hype. It shifts public transport from,
- 4:57you know, something that often loses money and needs subsidies
- 5:00to something that could actually be a profitable business.
- 5:02Exactly. Incredible.
- 5:03So if the money works, and it really looks like it does in
- 5:06theory, the next huge challenge has to be operational, right?
- 5:11The money doesn't matter if the cars are all stuck in the wrong
- 5:13part of town, or worse, they're all out of battery.
- 5:16How do you run these massive fleets efficiently?
- 5:18This is where we shift from economics to logistics, and the
- 5:22problem is really twofold. You have constant network
- 5:25imbalance. Meaning the cars end up where
- 5:28the people aren't. Exactly.
- 5:29They drop people off and then they just accumulate in low
- 5:31demand areas leaving the busy spots with no cars.
- 5:35And the second problem is proactive charging.
- 5:38Right, You can't have your whole fleet decide it needs to charge
- 5:41at 5:00 PM when everyone is trying to get home.
- 5:43If they only charge reactively when the battery hits say 5%,
- 5:49they'll all hit that limit at the same time, right during peak
- 5:52demand. And the system collapses.
- 5:54It would, and This is why the controlled algorithms are so
- 5:56important. Our sources detailed this
- 5:58predictive framework called Tea Mod.
- 6:01Tea mod. What does that stand for?
- 6:03Tunable electric autonomous mobility on demand.
- 6:06But you can just think of it as like the air traffic control for
- 6:09a fleet of robo taxis. I like that it's not just
- 6:12telling cars where to go, it's predicting where they'll need to
- 6:15be. Exactly.
- 6:16It's designed to manage the chaos before it happens.
- 6:19It does this in three interconnected steps.
- 6:21The 1st is dispatching. OK, so matching me with a car
- 6:25when I request a ride. Right, but it's not just about
- 6:28matching you with the closest car.
- 6:30It uses a mathematical tool ometimes called the Hungarian
- 6:33algorithm. Don't need the details on that.
- 6:35What's the output? What does it prioritize?
- 6:38It prioritizes minimizing your waiting time for one, but it
- 6:42also looks at the vehicles, state of charge and even
- 6:45something called a starting cost.
- 6:47A starting cost? What's that?
- 6:48It basically favors a vehicle that's already moving over one
- 6:51that's parked. Oh, that's clever.
- 6:53So if a taxi just drops someone off near me, it's better to send
- 6:56that one instead of waking up a parked car across town and
- 6:58having it drive over empty. Exactly, and minimizes that dead
- 7:02mileage which phase energy and cuts wait times.
- 7:05The second step is rebalancing. And this is where the predictive
- 7:08part really comes in. Yes, using a technique called
- 7:11Model Predictive Control or MPC. So MPC, It sounds like something
- 7:16I don't know NASA uses to land a rocket.
- 7:19How does this futuristic model manage a bunch of empty cars?
- 7:22Well, it manages them by thinking ahead instead of just
- 7:25reacting to what's happening now.
- 7:27Oh, all the cars are downtown. Let's send some Uptown.
- 7:30Right. Instead of that, it uses
- 7:32forecast and future demand. The system repositions empty
- 7:35cars based on where it predicts the imbalance will be in the
- 7:38next hour. So it's playing chess, not
- 7:40checkers. It knows the big concert ends at
- 7:4310:00 PM, so at 945 it starts moving empty cars toward the
- 7:47stadium. That is the perfect analogy, and
- 7:49that predictive insight is what makes the final steps smart
- 7:53charging work so well. OK.
- 7:55So it embeds the charging decision into that same
- 7:57predictive model. Exactly.
- 8:00Traditionally charging is an emergency.
- 8:01Your car is about to die so you have to stop.
- 8:04TE mod she treats it differently.
- 8:06How so? The goal isn't just to charge
- 8:08the car, it's to charge the car without impacting service.
- 8:11The system finds the natural downtime in the city's rhythm.
- 8:14Like between 11:00 AM and 2:00 PM.
- 8:17Or overnight, and it proactively sends cars to charge during
- 8:20those low demand windows. So it forces the charging to
- 8:22happen when nobody needs the cars anyway.
- 8:24Which means you have maximum vehicle availability during the
- 8:27morning and evening rush hours. The system treats charging as a
- 8:31resource to be managed, not a crisis to be solved.
- 8:34And the results show this actually works.
- 8:36It reduces wait times. Significantly compared to just
- 8:40reactive charging, this predictive approach allows the
- 8:43fleet to maintain a really high level of service even under
- 8:47pressure. So if the economics and the
- 8:50operations are tilting so heavily toward these big shared
- 8:54fleets, it brings up a big question for us as individuals.
- 8:58The ownership dilemma. Right.
- 9:00Should I own a vehicle at all, even an automated one, or just
- 9:03use these services and? To answer that, you have to look
- 9:05beyond just the financial costs. You have to factor in the non
- 9:08financial benefit time. The time you get back because
- 9:11you're not actively driving, how do you even put a price on that?
- 9:14Well, we have some data. In the UK, for example, a
- 9:17typical driver spends an average of 274 hours a year just.
- 9:22Driving 274 hours, that's almost 7 full work weeks.
- 9:26It is, and that time has a real financial value.
- 9:29The estimated productivity benefit is valued at about £20
- 9:32per hour. So you're not just saving the
- 9:34cost of the car, you're gaining thousands of pounds worth of
- 9:37productive time back. For a lot of people, the ability
- 9:40to work or sleep or just relax for those 274 hours is worth
- 9:45more than the cost of the service.
- 9:47And Speaking of that future, we're already seeing some pretty
- 9:50radical proposals out there. I'm thinking of what Elon Musk
- 9:52has talked about with the cyber cab and the robo van.
- 9:55Right, these are not just automated versions of today's
- 9:58cars. Not at all.
- 9:59The Cyber cab is described as this futuristic 2 seater
- 10:02vehicle. No steering wheel, no pedals,
- 10:05just a pure shared robo taxi. With a theoretical price that's
- 10:09actually pretty low, around €27,000 and a target date of
- 10:132027. And this really gets at the the
- 10:16big technological split in the industry, doesn't it, how
- 10:19different companies are trying to solve this problem?
- 10:22Yeah, they're basically two big philosophies here.
- 10:24You've got the Tesla approach. It's very software centric.
- 10:27All about AI, machine learning and using that massive amount of
- 10:31data from the cameras on cars they've already sold.
- 10:33Exactly. It's a vision first path, yeah.
- 10:35Then on the other side you have companies like Waymo.
- 10:38The hardware heavy approach. The hardware heavy approach,
- 10:40yeah, using LIDAR, that laser radar to build these super
- 10:44precise 3D maps of the world around the car.
- 10:47It's different bet. But no matter which tech path
- 10:50wins, they both promise these incredibly low operational costs
- 10:54once you take the driver out. Absolutely.
- 10:56The projection is startlingly low, something like 30 to $0.40
- 11:00per mile, but. There's always a but.
- 11:02There's always a But the promise of this cheap 24/7 future has to
- 11:07be measured against the messy reality of actually putting it
- 11:09on the streets, which brings. US to the real world resistance.
- 11:12Yes. Consider this one anecdote from
- 11:15Turin in Italy. It's a perfect example.
- 11:18A manager broke his ankle and he couldn't find a single taxi to
- 11:21take him on a short trip. Even during the cities tech week
- 11:24of all times. It just highlights a basic
- 11:26failure in the traditional taxi system.
- 11:28They often don't want to take those short, inconvenient trips.
- 11:31And those traditional systems have a history of pretty
- 11:36aggressive resistance to competition, like Uber.
- 11:39So the conflict is set. On one side.
- 11:40You have this new technology promising cheap, accessible 247
- 11:45mobility for everyone. And on the other.
- 11:47You have these entrenched industries and the governments
- 11:49that regulate them, and they might react by creating what
- 11:53some sources called absurd norms.
- 11:56Regulatory hurdles designed not for safety, but just to protect
- 11:59the old way of doing things. And that could mean condemning
- 12:02customers like that manager in Turin to poor service.
- 12:06Even when a much better technology is ready and waiting,
- 12:09the cost of that resistance is paid by the consumer.
- 12:12So when we put this all together, what does it mean for
- 12:15you, the listener? We've gone from a shocking 75%
- 12:18cost drop all the way to these complex TEA mod algorithms.
- 12:21I think a core insight is pretty clear.
- 12:23The economics are just undeniably tilting toward
- 12:26autonomous electric fleets. Because you take out that huge
- 12:28labor component. You removed nearly half the
- 12:31operating cost. And on top of that, these
- 12:34sophisticated control policies show that the operational
- 12:36challenges like charging and balancing, they're solvable with
- 12:40predictive tech. The system isn't waiting for a
- 12:43crisis, it's preventing it. You know if we wrap this all up
- 12:46with one last thought for you to chew on, we focus so much on the
- 12:50direct costs, the TCO of the robo taxi.
- 12:54But when you think about that story from Turin, that stranded
- 12:58traveler, those 274 hours of lost productivity a year, yeah,
- 13:02maybe the most important TCO here isn't the total cost of
- 13:05ownership of the vehicle. Maybe it's the total cost of not
- 13:07having these services. The.
- 13:08Cost of inertia. The cost of inertia that might
- 13:11be the real measure of disruption here, the time
- 13:13wasted, the missed opportunities, all because we're
- 13:16protecting a system that doesn't always work.
- 13:19A powerful thought When the tech offers such huge gains in time
- 13:22and convenience, protecting the status quo really does become
- 13:26the most expensive option. Thank you for joining us for
- 13:28this deep dive into the world of robotaxis.
- 13:31If you found this discussion valuable, please follow us
- 13:34wherever you listen to your deep dives to make sure you catch our
- 13:36next episode. We look forward to diving in
- 13:39with you again soon.