- The Driverless Digest
- Posts
- Waymo is coming..
Waymo is coming..
But they drive slower than an Uber. Cruise makes a comeback (sort of), and Wayve launches.
Top Stories of the Week
How San Francisco Learned to Love Self-Driving Cars (link, no paywall). I posted on LinkedIn about this article but my big takeaway was that people really love Waymo for the peace and quiet of not having a driver.
When I first started driving for Uber and Lyft in 2014, passengers were eager to learn about their drivers and there was a lot of conversation and banter. These days, most riders like to get in the car, not talk, stare at their phones, and not leave a tip. Waymo is great for that.
A recent experiment by two reporters from the San Francisco Chronicle showed that Waymo takes different routes and drives slower, leading to longer trips, when compared to a human-driven Uber (link, no paywall). This was a neat and well-presented experiment that lines up with my own personal experience (albeit anecdotal) and makes sense since the Waymo Driver doesn’t go on freeways yet and follows the rules of the law more closely (ie stopping for yellow lights; no Uber driver in the history of the world has ever done this).
I’m confident that Waymo will start driving on freeways soon but the safety angle is an interesting tradeoff. Of course people say they want a safer ride, but more often than not, rideshare passengers are late, in a hurry, etc and there’s not much a Waymo can do to speed up compared to a human.
The other issue I’ve noticed recently with Waymo is with the pick-up/drop-off for tricky locations. On the way to our mobility happy hour last week, I took a Waymo but it wanted to loop around the block, make a U-turn and then drop me off a 3 minute walk away from the venue. Probably would have added 5-7 minutes to the trip if I didn’t hop out at a red light. An Uber driver would have done a much better job. I’m not sure how easily this can be fixed. Check out my video reenactment here.
Announcements | Partnerships | Funding
Tesla releases a demo video of Robotaxi wireless charging. This is pretty neat, and also a good idea. I invested in a similar company called Volterio that does just this.. Elon, hmu!
Robotaxi wireless charging
No hands required
— Tesla (@Tesla)
3:42 PM • Oct 18, 2024
Wayve Launches in the US with San Francisco Testing Program and New Office (link).
Waymo introduces Live Activities, a feature that allows you to track your trip, anywhere on your device, even on your lockscreen (link).
GM’s Cruise has removed humans from the driver's seat in some Houston tests—and moved them to the passenger seat (link, no paywall). Don’t call it a comeback…
Oh, but Cruise also lost $435M in Q3 2024 (link).
Mapless AI and Aero Corporation Partner to Pilot Tele-Operations for U.S. Car Rental (link)
First Ride
I recently had the chance to try out Waymo’s autonomous taxi in San Francisco, and I have to say—I’m totally impressed! (link)
AVs/Humans behaving badly
Edge cases, am I right? Waymo takes on an erupting fire hydrant (link)
“First time in a Waymo with trash. I’m livid” (link). Having a driver in the car prevents a lot of these types of issues but even then, some passengers are disrespectful even with a human in the car. At scale, this is going to be a huge issue as there’s a lot of potential fraud (yes vomit fraud is a real thing), he said/she said, and other issues that will need to be mediated by Waymo (expensive/not efficient). Especially as this moves form early adopters/fans to wide adoption, people unfortunately are going to treat these cars like crap.
Other Stuff
Waymos are coming…(link)
I thought this might be photoshopped but apparently it’s real. In order to help passengers find their vehicle, Waymo shows your initials and a custom color on the bright 360° display mounted on the top of the car. You can also honk the horn from the app. And you’re the only one able to unlock the vehicle from the app also.
The trickiest pick-ups and drop-offs happen at large events where there’s massive demand for rideshare. To this day, Uber and Lyft still haven’t figured out how to best handle these situations - it’s a thorny problem. Savvy passengers know the easiest way to get a ride is to walk a few blocks away from the venue and call a ride away from the ‘designated pick-up’ location.
But I feel like Waymos are going to struggle with drunk passengers all looking for their car at the same time/place since the vehicles all look identical. Add in GPS that doesn’t work well in downtown areas with skyscrapers or when thousands of people are on their phones in a constrained area and you may have some issues. Waymo’s 5 minute timer to get to the car is strict too.
You know what's more fascinating than 8,003 minutes spent in a Waymo? (link) It’s 28,641 minutes spent in a Waymo (link). Although there’s a slight asterisk since Daylen works for Waymo :)
Daily reminder that the only reason you can’t take a much safer self-driving car home from SFO is because of obstructionist local politics (link). Doesn’t make much sense to prevent Waymos from doing pick-up/drop-offs at SFO or LAX. Make them go to the LAX-it nightmare of a lot just like everyone else!
Waymo vehicles are empty about 55% of the time (nice find by @SafeStreetRebel). That’s not good, especially compared to Uber/Lyft at 20% in SF! This is what we refer to in the industry as deadhead time and it’s the worst since drivers (or Waymo) are not paid for this time. Waymo’s current fleet is supply constrained so I imagine their idle time (Period 1, waiting for a request) is low. But every time you see a 20 min. ETA instead of a 3 min. ETA (Period 2, driving to a passenger), that hurts the utilization rate big time.
But don’t get it twisted: Uber, Lyft and even Waymo are not the root cause of congestion in your city. They are an easy scapegoat though. Typically, Uber and Lyft make up less than 15% of the total VMT in dense cities like San Francisco. And as low as 3% in cities like Seattle. Math is fun.
Until next week.
-Harry