Travis Kalanick Thinks Uber Screwed Up on Robotaxis 😳

Waymo launches in Silicon Valley, seems crazy that AVs still don’t have autonomous charging, and why you shouldn’t mistake ridehailing for AV ridehailing

My annual Curbivore conference is coming up on April 11th in Downtown LA! We’ve got a great lineup of speakers and partners (Uber, DoorDash, Zoox, Waymo and more!) and AVs/automation/AI will all be big themes at the event. Plus, it’s held at a unique outdoor/indoor venue in sunny Los Angeles - unlike any conference out there, guaranteed!

Prices will be jumping up on Monday and TDD readers can get 50% off early bird tickets until then 🎉 (Public sector employees get in free!) 🎟️ See you there!

Top Stories of the Week

Travis Kalanick thinks Uber screwed up: “Wish we had an autonomous ride-sharing product” (link). Well that was quite the headline, but I’m not so sure I agree with TK here.

• Uber Advanced Technologies Group (ATG) was the first and only AV company to kill a bystander.

• They invested billions of dollars into self-driving before the accident, and would have had to spend billions more. I’m no financial expert but I’m guessing Uber’s public market investors might have had an issue with that.

• Self-driving technology was not a sure thing like it is today with Waymo. There was no guarantee that it would work this well even if you invested billions of dollars more.

• TK was also at the helm of the whole fiasco with Anthony Levandowski.

I think Uber is now doing what they do best, generating a ton of demand and building a framework for partners like Waymo (and Zoox, May Mobility, etc in the future) to plug into their network seamlessly. And then the Waymos of the world can focus on the technology side of things and make their product as great as possible, while reducing the cost, and expanding the Total Addressable Market (TAM) to every personally owned vehicle in the world.

Image Credits: Bloomberg/Contributor/Getty Images

Waymo expands its robotaxi service across Silicon Valley (link). They're now offering public robotaxi rides to folks in its "Waymo One" early rider program across Mountain View, Los Altos, Palo Alto, and parts of Sunnyvale. Waymo is doing a great job with weekly announcements and continuing to build buzz and momentum. But part of me is starting to wonder if they’re moving fast enough.

Waymo doesn’t have any real competition in the United States, but if AV tech is eventually commoditized, they don’t have a strong moat or network effects, unless they are ubiquitous. First mover advantage is real in the rideshare industry, and it’s the reason why Lyft will never disrupt Uber even though it’s the same drivers, product, result, etc. So if I can get an AV on Uber (or Lyft) from Zoox, May Mobility or even Apollo Go, isn’t that just as good as hailing a Waymo?

Cool Rides

Finally found a Waymo on the Uber app in Austin (link). This is the big complaint I’ve heard from riders so far in Austin - it’s impossible to actually get a Waymo on the Uber app. Customers need to be signed up for ‘Waymo on Uber’, request a regular UberX ride, and then hope for a Waymo. Frustrating for customers who want a Waymo, but I think the product strategy is smart since Uber is known for reliability. Every time you open the app, you have a car at your fingertips and a 3-5 minute ETA. Now you may not like the car, or the driver, or the price, but it will get you from A to B.

Uber could add a ‘Waymo’ option to the app but since there’s limited supply, you would still need to have a human driver back up. I can’t imagine Uber would ever allow their app to show ‘no cars available’ or a 20 min + ETA like we see all the time on Waymo here in Los Angeles. So even though some customers are disappointed they’re not getting a Waymo, that’s actually a good thing for both companies since it means Waymo vehicles are at max utilization. And then Uber still gets to serve the customer a ride with a human driver, instead of the ride not being completed (like what would happen on the Waymo app).

Took my first ride in a driverless car last night, and I have to say—I’m impressed. 🚗🤖 (link).

I took a self-driving Waymo car in Phoenix today. My skepticism lasted about 30 seconds and then I was used to it (link).

AV behaving badly

Waymo almost causes accident (link). ‘Waymo could have caused an accident’ would have been a better title for this video. That being said, this does seem dangerous and I would not have wanted to be in the back of that Waymo as it sat in the intersection trying to make an unprotected left turn lane on a busy 3 lane street. Feel like these videos are few and far between so that’s a good thing.

Other Stuff

Seems crazy that we still don’t have autonomous charging…everything else is autonomous but a human plugs in the charger…feels solvable (link). I invested in a company (Volterio) that does this so I couldn’t agree more. I like the idea of automating binary tasks like this - plug in/take out. Automating other stuff like cleaning is tougher since there’s so many different scenarios and edge cases (trash, smudges, wet stains, puke, etc). The cottage industry around fleet maintenance is going to be huge though.

👏 Amazing tote and marketing. I’ve been enjoying the SF Standard’s reporting on AVs and more broadly. Worth taking a look.

Can AVs Make Ride-Hailing more affordable? (link). Gridwise Analytics and S&P Global dropped a whitepaper diving into this question. And if you enjoyed this research, Gridwise will also be giving a keynote at Curbivore that should cover some exciting data related topics.

Don’t Mistake Ridehailing for AV Ridehailing (link). Nice article by Reilly Brennan here and echoes a lot of similar thoughts I’ve shared in the past but with better graphics :) At peak demand, you can also charge peak pricing across every trip, so the potential revenue of these trips ($ per trip) is more valuable than at baseline. So if Waymo can’t serve those peak trips, it’s a poor rider experience (no cars / high ETA) and a lot of missed revenue (especially since Waymo doesn’t seem to surge nearly as high as Uber does during peak demand).

When it comes to Tesla, the other big advantage they have is vertical integration. It would be smart for them to own the vehicles during baseline demand since they also manufacture the cars. Hertz has had a lot of success renting Model 3’s to Uber drivers but the cars took a massive depreciation hit as Tesla continually lowered the list prices over the past few years. Depreciation is one of the trickiest parts about owning a shared fleet. Get it wrong and you may go bankrupt.

How Uber’s new product chief is gearing up for robotaxis (link). I’m a big fan of Sachin Kansal and it must be an exciting time to work at Uber - there’s going to be a lot of fun/interesting work figuring out how to integrate AVs and humans across mobility and delivery. I can only imagine what the product road map looks like..

How Waymo helped turn the walk of shame into the ride of pride (link).

Ed Niedermeyer, co-host of the Autonocast podcast and the Ride AI conference (which I’ll be speaking at), did a Reddit AMA (link). I always like Ed’s analysis because he is in the weeds, technical and detail oriented.

Robotaxis without a brake pedal or mirrors? Not so fast, feds say (link, no paywall).

Pony.ai CEO on Tesla’s robotaxis: ‘We haven’t seen their large scale testing yet’ (link). The original headline of this article was ‘Pony AI CEO Doubts if Tesla Can Launch its Robotaxi Service Soon’ and boy did I get excited. But then I watched the interview and it turns out he didn’t say anything of the sort. And CNBC also later updated the title. The ironic part is that I think the original headline was reasonable. Tesla isn’t launching a robotaxi service comparable to Waymo anytime soon but they sure are fun to talk about.

Japan’s Nissan tests driverless vehicles in city streets filled with cars and people (link).

---

Enjoying The Driverless Digest? Share it with your friends and score rewards when they subscribe! For a limited time, refer five people and get a free ticket to Curbivore! 🚀🎟️

Until next week.

-Harry