Best of LinkedIn: Autonomous & Mobility Markets CW 41/ 42

00:00:00: Provided by Thomas Allgaier and Freeness, based on the most relevant LinkedIn posts about autonomous and mobility markets in calendar weeks forty-one and forty two.

00:00:08: Freenes supports tier one automotive suppliers with early stage market validation for their R&D efforts by combining secondary research direct OEM expert interviews and facilitated customer meetings.

00:00:19: you can find more info in the description.

00:00:22: Welcome to The Deep Dive!

00:00:23: Yeah so imagine for a second An entirely empty driverless car, just pulling up to your driveway To take you the airport.

00:00:32: I mean sounds super futuristic right?

00:00:33: Oh absolutely like write out of a sci-fi movie

00:00:36: Exactly.

00:00:37: But what if the real story here Like the actual multi billion dollar battleground isn't The AI software driving the car at all

00:00:45: Right?

00:00:45: What If the Real Story Is Actually The Massive Real Estate Empire The Underground Data Centers And Just The Physical Cleaning Crews Required?

00:00:52: Keep That One Car On The Road.

00:00:54: It completely flips how you view the industry, because for the longest time.

00:00:58: The only question that really mattered was well can this software accurately see a pedestrian in a crosswalk?

00:01:04: Right!

00:01:04: The basic cannon drive test...

00:01:06: Exactly and while the tech obviously isn't flawless yet we have officially crossed a threshold.

00:01:13: I mean the underlying question from mobility professionals right now is shifting.

00:01:16: it's now about Can these companies actually scale?

00:01:20: Can they integrate these vehicles into dense, chaotic cities and somehow turn a profit doing it?

00:01:26: Yeah.

00:01:26: And to really grasp that scale you just have to look at the massive geographical explosion happening across the globe over the last couple of weeks.

00:01:33: It's

00:01:33: everywhere!

00:01:35: We aren't talking about isolated carefully manicured test tracks in Arizona anymore.

00:01:40: This tech is hitting highly complex international markets.

00:01:43: So let's start with Waymo.

00:01:45: Okay yeah

00:01:46: There was this great observation circulating from Klaus Habermeyer, noting that Ramo has aggressively targeted Munich for its first German service and they have a goal of commercial deployment by late twenty-twenty seven.

00:01:58: Wow!

00:01:58: Twenty-Twenty Seven?

00:01:59: Yeah That's fast yeah And it doesn't even stop in Europe.

00:02:02: Lars Christian Grotem Olsen highlighted that Waymo is also expanding to Singapore for twenty twenty eight

00:02:08: Right which would be their very first Southeast Asian market.

00:02:11: Exactly

00:02:11: It's massive footprint expansion.

00:02:14: But you know if we connect us the bigger picture.

00:02:16: What's truly fascinating here isn't just where these cars are physically going.

00:02:20: What do you mean?

00:02:21: Well, it's how the underlying ownership models of these autonomous fleets are completely fracturing.

00:02:26: so to give you a perfect example Bojan Jukovic pointed out a really fundamental shift happening with Bolt.

00:02:33: Oh right!

00:02:33: The ride-hailing app.

00:02:34: Yeah bolt is going to outright own and operate twenty five thousand lucid robo taxis And these are equipped with NVIDIA Hyperion technology.

00:02:42: Wait...twenty five

00:02:43: THOUSAND?!

00:02:44: Yes They're buying them.

00:02:46: Yes,

00:02:46: it is a staggering move.

00:02:47: It's Boltz very first owned autonomous fleet in Europe?

00:02:52: It really signals that shared mobility operators realize they can't just provide a sleek app anymore.

00:02:56: They actually have to step up and physically own the heavy metal.

00:03:00: okay let's unpack this because This feels like a massive kind of counterintuitive pivot to

00:03:06: me.

00:03:06: how so

00:03:07: well.

00:03:07: traditionally The whole ride hail model his asset light.

00:03:11: Uber traditionally acts like Airbnb.

00:03:13: They just facilitate the transaction without actually owning the real estate.

00:03:16: Right, The drivers own cars Exactly

00:03:19: But Bolt is coming in here and actually buying entire apartment building so to speak.

00:03:24: Why would a software platform choose on that massive heavy capital risk?

00:03:30: That's great question.

00:03:31: I

00:03:31: mean why not just license the network And let someone else deal with car payments?

00:03:35: You'd think pure software plays smarter business move.

00:03:39: The reality on the ground is just far more complicated.

00:03:42: It all comes down to the regulatory environment in the European Union.

00:03:45: Oh, of

00:03:45: course!

00:03:46: Red tape... exactly.

00:03:47: The bar for homologation basically getting a vehicle legally certified for the road is incredibly high there In Europe, the fleet operations, the autonomous software and physical car must work together as one indivisible system

00:04:01: I see.

00:04:02: So you can't point fingers if something goes wrong.

00:04:04: Precisely you cannot fragment the responsibility.

00:04:07: If a sensor fails or, y'know...the software glitches.

00:04:10: European regulators won't tolerate Bolt blaming Lucid or Lucid blaming NVIDIA.

00:04:15: It's all in Bolt

00:04:16: Right.

00:04:17: By owning the fleet, Bolt controls maintenance The sensor calibration, the software updates literally everything.

00:04:23: And they already have street teams, depots and charging infrastructure from their shared scooter & car businesses.

00:04:29: That makes total sense.

00:04:30: when you frame it around liability You essentially have to own the whole puzzle to guarantee the picture.

00:04:34: Exactly!

00:04:35: In this race-to scale operations, it isn't just a US and Europe story either.

00:04:40: The Middle East is moving incredibly fast.

00:04:43: Oh

00:04:43: absolutely

00:04:44: Yeah.

00:04:44: Sergei Kochnev noted that fully driverless Baidu Apollo Go And we ride.

00:04:48: robo taxis are actually already running via the Uber app in Dubai.

00:04:52: Wow...and

00:04:53: on top of that Nassir Alsari mentioned that Memento is planning to scale thousands of robo taxes in that exact same market by twenty twenty seven.

00:05:01: Yeah, Dubai is aggressively positioning itself as the ultimate global test bed for this exact reason.

00:05:07: I mean they have a stated government target where twenty-five percent of all transport must be autonomous by twenty thirty.

00:05:14: Twenty five percent?

00:05:15: Right!

00:05:16: That's huge!

00:05:18: So instead fighting the operators The Government is actively offering digital infrastructure High definition mapping environments and clear liability frameworks that these companies desperately need.

00:05:29: They needed to prove their tech works at scale.

00:05:31: Exactly!

00:05:32: Which perfectly brings us the biggest name in the EV space, because we just discussed how operators like Bolt are taking on massive capital risk and Europe.

00:05:40: due regulation.

00:05:41: Right

00:05:42: But In

00:05:42: U.S.,

00:05:43: The regulatory environment is a bit more fragmented And that allows for completely different approach.

00:05:49: So let's talk about Tesla Cybercap.

00:05:52: Yes...the big reveal.

00:05:54: Maxim Gleger recently shared his hands-on experience with a cyber cab roll out in Austin, and the vehicle is striking.

00:06:01: I mean no steering wheel No pedals

00:06:04: totally stripped down yeah

00:06:06: And it's priced competitively per mile.

00:06:07: with an uber X Plus.

00:06:09: The car itself has targeted to be incredibly cheap to manufacture supposedly well under thirty thousand dollars.

00:06:15: And here's where Tesla strategy diverges wildly from what Waymo or Bolt are doing.

00:06:21: Atlas-Barry cut a very critical detail on Tesla's web form for the CyberCab.

00:06:27: Oh, well was it?

00:06:29: That form strongly signals a highly asset light model for Tesla.

00:06:33: Interesting.

00:06:34: So the intent seems to be that outside individuals and companies will actually buy these fleets.

00:06:39: Oh, interesting.

00:06:40: Yeah so they manage the massive charging depots handle The Daily Cleaning and Tesla simply acts as a software platform taking a revenue cut of every ride.

00:06:50: So it's exactly opposite to both.

00:06:51: strategy in Europe Exactly.

00:06:54: And we are seeing early signs this hardware moving onto streets quickly.

00:06:57: Steven Snyder analyzed Texas DMV data and found that registered Texas cybercabs jumped from just forty five To one hundred twenty six In span of twenty two days.

00:07:07: But Okay, here is where I want to push back on Tesla's immediate operational reality.

00:07:13: Milena Boychev noticed a very curious... Honestly kind of funny... ...operational quirk during this Austin rollout.

00:07:20: What was that?

00:07:22: The cyber cab apparently has a strict eleven p.m curfew.

00:07:25: Oh yes!

00:07:26: Now the curfee.

00:07:27: Wait

00:07:27: a minute.

00:07:27: Tesla vision stack is supposedly this incredibly sophisticated system That preserves extreme detail in pitch darkness right?

00:07:34: That's the claim.

00:07:35: So if the cameras and AI are really that good Why send the cars home at eleven

00:07:39: p.m.?

00:07:40: Are they afraid of the dark?

00:07:42: Or are they just afraid to this sudden spike in highly unpredictable impaired pedestrians and reckless human drivers after

00:07:47: last

00:07:47: call?".

00:07:48: It's not that cameras can't see the drunk pedestrian stumbling out a bar, it is that managing chaotic irrational physics of late night behavior requires totally different levels risk tolerance.

00:08:02: This is the classic example of strictly limiting a vehicle's operational design domain, or ODD.

00:08:08: Okay for anyone not deep in the engineering.

00:08:10: weeds translate Operational Design Domain for

00:08:14: us.

00:08:15: Sure think of the ODD as an invisible digital fence.

00:08:19: it defines exactly where when and under what conditions The autonomous system is certified to operate safely

00:08:26: like specific weather or speeds.

00:08:28: Exactly It might mean the car can only drive in specific neighborhoods, only under forty-five miles per hour and when it's not raining.

00:08:36: By implementing an eleven p.m curfew Tesla is essentially saying our ODD currently excludes high liability chaos of midnight traffic.

00:08:45: So they are actively geofencing out the unpredictability.

00:08:50: Exactly because liability is the single biggest friction point right now, especially when you look overseas.

00:08:57: For instance Sophie Green highlighted a recent Belgian test where Tesla running FSD supervised was caught speeding and illegally overtaking cyclists in residential zones.

00:09:08: Wow that's not good luck.

00:09:10: No it isn't.

00:09:12: And it exposes massive dangerous paradox with level two systems because it's technically an assist system and not fully autonomous.

00:09:21: The human driver sitting in the seat still carries one hundred percent of the legal liability for those traffic violations, Not Tesla!

00:09:29: Right so...the car makes a decision to break the law but you are the one who pays the ticket and gets the points on your license.

00:09:36: Exactly That is a massive legal gray area

00:09:41: It Is But regulators are trying to find some middle ground.

00:09:45: Jean Drogaba reported that Germany is actually backing swift EU-wide approval for Tesla system, but they reached a fascinating compromise to get there.

00:09:53: What's the compromise?

00:09:54: They agreed to let Tesla roll it out provided they simply rename it to Tesla assisted driving just To explicitly clarify to the driver that they are still legally in charge.

00:10:05: It's amazing what little rebranding can accomplish and the eyes of law right

00:10:09: truly.

00:10:11: By the way, if you're finding these insides valuable for your own strategic planning make sure to hit subscribe so that you don't miss our future deep dives.

00:10:18: We love digging into operational realities with you!

00:10:21: We really do and operations are at the core of this issue here.

00:10:25: Tesla wanting outside fleet owners buy their cybercaps leads us directly in the brutal behind-the scenes math running a robo taxi business because

00:10:33: owning car is actually the cheap part.

00:10:35: Exactly

00:10:36: The math really is wild.

00:10:37: Let's look at the theoretical upside first, just to see why so much money is pouring in.

00:10:42: Chamath Palihapatia shared some interesting napkin mouth on this recently... Are

00:10:47: the profitability projections?

00:10:48: Yeah!

00:10:49: ...the projection is that a RoboTaxi operating in a dense high-demand urban area could drive fifty thousand to one hundred thousand miles per year.

00:10:58: After basic costs That vehicle could theoretically net roughly thirty thousand dollars annually in profit Per Vehicle.

00:11:07: So when you look at those numbers on a spreadsheet, You can see why investors are salivating.

00:11:10: It looks like a money printing machine.

00:11:12: Well it looks great on a spread sheet assuming the vehicle exists in a sterile vacuum.

00:11:16: But we need a serious reality check on the actual logistics.

00:11:20: bring us back down to earth.

00:11:22: Jin Hua Zhao interviewed the CEO of Move AV, which is a company that actually operates Waymo fleets in several US cities.

00:11:29: and The most startling fact from that conversation Is that the autonomous vehicle itself?

00:11:34: Hardware the expensive sensors the AI compute only accounts for thirty two forty percent Of the five-year operating cost.

00:11:42: wait really yeah Only thirty to forty percent.

00:11:46: The other sixty to seventy percent of the cost is entirely consumed by the massive operational ecosystem required just to keep that car on the road.

00:11:54: Okay, so if I'm understanding this right... ...the self-driving cars are just the razor and physical fleet operations or the razor blades?

00:12:03: That's perfect analogy!

00:12:06: Jeff Teisman made a brilliant point along these lines.

00:12:09: He noted that software can Thread through density traffic beautifully.

00:12:14: It can yield a pedestrians and navigate roundabouts, but you know what software doesn't do?

00:12:19: What's that?

00:12:20: it does not wipe down the backseat after a messy passenger spills of coffee.

00:12:24: Nope

00:12:25: Software doesn't inspect the tire tread And it certainly doesn't physically plug the car into a fast charger.

00:12:30: A robot taxi only earns that theoretical revenue when he is clean charged and Staged in the exact right location at the exact bright time.

00:12:38: Right And scaling that operational razor blade ecosystem requires massive physical infrastructure, and usually in some of the most expensive real estate markets in the world.

00:12:49: Like downtown San Francisco or Manhattan?

00:12:52: Exactly.

00:12:53: Carson Jones warned The Industry That Robotaxi Growth Isn't Just A Story About Manufacturing Cars.

00:12:59: It's a story about acquiring prime land securing massive electrical power grids building cooling systems.

00:13:08: These regional charging depots are essentially giant data centers on wheels,

00:13:13: right?

00:13:13: Because if a car gets dirty on a Friday night at nine p.m.. It has to take itself out of service drive five miles across town To a Depot get manually cleaned by human crew and then drive back.

00:13:25: The opportunity cost of that downtime is huge.

00:13:27: It really is,

00:13:28: and brings up another often ignored financial metric that Eduardo Rojas highlighted.

00:13:33: this metric will quietly determine who actually wins the total costs of ownership

00:13:38: war.

00:13:38: let me guess vehicle depreciation.

00:13:41: bingo.

00:13:42: oh This is a fascinating point.

00:13:44: Let's dig into this.

00:13:46: Normally A rental car company like Hertz buys a fleet uses them for three years And then sells them on the secondary market to recoup a massive chunk their capital.

00:13:54: right

00:13:55: Exactly, that's the standard model.

00:13:57: But if Bolt or a Tesla fleet owner buys thousands of purpose-built robo taxis without steering wheels and pedals what is the resale value for that pod in five years?

00:14:09: You can't exactly sell it on the used market to a teenager looking their first car!

00:14:14: You definitely cannot.

00:14:16: The secondary market for a steering wheelless vehicle currently does not exist.

00:14:22: The residual value at the end of its life might literally just be the scrap weight of it's battery pack.

00:14:27: Just

00:14:27: scrap

00:14:27: metal?

00:14:28: Yeah If you can't recoup capital on the back-end, your upfront operations have to be incredibly lean and highly utilized just to survive.

00:14:38: But I mean if they can solve that brutal operational math... ...the potential impact on society is staggering.

00:14:44: Chris Martinez shared an estimate from Ark Invest suggesting there are two point four million.

00:14:50: highly utilize robo taxis could theoretically cover the vast majority of all urban miles drifted in the US.

00:14:57: Two point four million!

00:14:58: That's a tiny fraction if what is out there now?

00:15:00: Exactly

00:15:01: Think about that.

00:15:02: We currently have over two hundred and fifty million personal cars in the U S And they set parked and useless.

00:15:08: ninety five percent at time.

00:15:10: Twenty-four million active robo taxis Could fundamentally reshape The entire model car ownership.

00:15:16: But, that projection immediately forces us to look At the Urban Environment Because if we are about to drop millions of high utilization robotaxes onto city streets, how will our physical cities actually handle them?

00:15:28: Good point.

00:15:29: We have to move from the business economics to ground level urban realities

00:15:33: Right like actual impact on the curb.

00:15:36: Matthew Raithman posted a really interesting strategic move by Waymo in San Francisco.

00:15:41: What do they do?

00:15:41: They're

00:15:42: actively offering a two dollar and eighty five cent credit for waymo rides that were taken within two hours of public transit trip.

00:15:49: Oh, that's smart.

00:15:50: It is and it signals a massive shift in the conversation.

00:15:54: The debate he's clearly moving past.

00:15:56: just Is the car safe to?

00:15:59: how does this fleet integrate with existing public transit?

00:16:02: And How does it affect our curb management?

00:16:04: an congestion is the operative word That city planners are truly terrified of.

00:16:11: Clebber Sally raised A major warning flag about what He calls empty repositioning trips

00:16:16: right the ghost cars.

00:16:17: Exactly, think about it.

00:16:19: If a robo-taxi drops someone off downtown in the morning rush hour but then next ride request is five miles away out on the suburbs that car has to drive empty to get there

00:16:28: Just adding into traffic

00:16:29: Right!

00:16:30: if we aren't careful We risk simply automating traffic congestion rather than solving It...we could just be filling roads with empty ghost cars driving in circles waiting for a ping

00:16:40: Which really makes me question The core strategy of this entire industry Especially depending where you are In the world.

00:16:46: Daniel Abreu Marquez analyzed the global landscape and argued there are actually three distinct robo-taxi races happening simultaneously.

00:16:55: Three races, break that down!

00:16:57: So in the

00:16:57: U.S.,

00:16:58: robo taxis are essentially being deployed as a substitute for public transit because frankly US Public Transit is often lacking.

00:17:06: True

00:17:08: In China it's massive topdown industrial policy push to dominate future of mobility.

00:17:14: but Europe The goal is to use them as an extension of public transport, like a first mile last-mile solution.

00:17:20: Okay that makes sense

00:17:22: But wait let's look at the actual data.

00:17:25: Simon Cox pointed out That the vast majority Of all robo taxi trips carry two or fewer passengers

00:17:31: Right mostly single riders.

00:17:33: Yeah.

00:17:33: so if it really just one Or Two people per trip isn't focusing on individual autonomous cars?

00:17:38: The completely wrong approach for Europe's density.

00:17:42: I mean, shouldn't we be obsessing over autonomous buses and high-capacity transit instead of clogging the streets with more two seater pods?

00:17:49: That is the exact tension that European urban planners are wrestling with right now.

00:17:54: An autonomous car is still a physical car.

00:17:56: It's takes up fifteen feet of asphalt whether human driving it or computer space efficiency critical in medieval European city grid.

00:18:05: You can just expand roads.

00:18:06: No you cant.

00:18:08: But beyond the spatial geometry of the streets, there is a much deeper fundamental barrier we haven't really solved yet which is public trust.

00:18:16: Right do people actually want to get in them?

00:18:19: Saber Fala pointed out his street interview conducted recently by ITV and London.

00:18:23: when asked almost everyone interviewed point blank said they would not trust getting into a self-driving taxi

00:18:29: Really even though statistically less likely to get into a crash than with the human driver?

00:18:35: Exactly.

00:18:36: The technology might be statistically safer on a spreadsheet, but psychological trust isn't built on statistics.

00:18:42: it's build on perceived control and the trust simply isn't there yet for the average commuter.

00:18:48: So if public trust is really that low I'd imagine local governments have to step in and basically force A sense of accountability.

00:18:57: How are regulators responding to that lack of trust like how do we bridge that gap?

00:19:02: through severe financial and legal accountability.

00:19:05: Brian Reimer shared an excellent example of this playing out right now, California just passed SB-twelve forty six.

00:19:11: What does that do?

00:19:12: This is a new law that actively finds robotaxi operators up to ten thousand dollars if their autonomous vehicles obstruct first responders like fire trucks or ambulances during an emergency.

00:19:22: Wow!

00:19:23: Ten thousand dollars per incident.

00:19:25: Yeah

00:19:26: It's very clear hardline signal from regulators.

00:19:29: The AI might be doing the driving, but a multi-billion dollar corporation is holding the license.

00:19:34: And that organization must be held financially accountable when the real world gets messy right?

00:19:39: It's no longer enough that your car can technically drive from point A to Point B. Your company has to be able to manage the chaos When a fire breaks out on point A.

00:19:49: That it such a vital point and don't...the robot is turning in wheel But the corporation asked answer of this city.

00:19:54: We've covered incredible ground today, really exploring how autonomous mobility is maturing at lightning speed.

00:20:11: New

00:20:14: episodes drop every two weeks.

00:20:17: Also, check out our other editions on charging and battery tech commercial fleets and SDV and AI defined.

00:20:24: make sure to subscribe.

00:20:25: yeah please do.

00:20:26: And we want to leave you with a final thought to ponder as You look ahead at this rapidly shifting mobility landscaping maybe easy walkout your own driveway tomorrow morning if robo taxis eventually cost half As much per mile is owning an ensuring of personal car And the software inevitably learns to drive much safer than we easily distracted humans do.

00:20:45: Will the ultimate status symbol of future be owning a vehicle that actually still has steering wheel?

00:20:51: That is fascinating question, thanks for diving deep with us!