This week Zoox, the autonomous vehicle company Amazon recently acquired, unveiled an autonomous electric vehicle that they supposedly plan to make available as part of a Robotaxi service. Conceptually, it’s fairly similar to the Cruise Origin, another AEV with a bidirectional design:
There are a few things to like here. The idea of an autonomous-only vehicle that has no front or back and can travel in either direction seems like a winner, and I expect vehicle design to follow this route over time as autonomous software becomes more reliable. It’s also great that the vehicle is pure electric, and designed from the ground up for autonomy –– a stark contrast to Waymo’s approach of retrofitting existing, battle-tested vehicles from major legacy automakers.
Of course, there are some big buts here. Everyone wants to do an autonomous vehicle design with no steering wheel, but that kind of hardware design sets a very high bar for the software to leap over. It’s very difficult to launch autonomous software for the first time and have it do everything perfectly. Problems, unexpected situations, and bizarre corner cases are almost guaranteed. Many iterations will be needed to improve reliability and comfort. That’s why I strongly believe there will be a transitionary period where self-driving vehicles have a steering wheel and pedals to allow for training feedback and manual override of the computer’s driving. Without these controls, riding in an early autonomous vehicle will be pretty annoying.
The other problem with this concept is that the vehicle will be too expensive for people to buy, so Zoox isn’t even going to try and sell it to consumers. If you want to ride in one of these vehicles, you’ll have to order a ride just like you would with an Uber. This kind of business model will become increasingly popular as autonomy advances, but it means that Zoox won’t be able to fully replace car ownership. If you want to take a road trip, keep your stuff in the car, or have your vehicle ready whenever you need it… Zoox is not for you.
At least initially, I don’t think consumers will enjoy sitting powerlessly in the back seat of a Robotaxi hoping they don’t die. Until people are extremely comfortable with the reliability of a given autonomous software package, they’re going to want to be in control. In my view, most of these ground up Robotaxis are still a few years ahead of their time. Without something people can start using now, players like Zoox risk being forgotten as consumers try offerings from Tesla, Waymo, and Cruise.
Personally, I would also be scared to be facing backwards while an autonomous car drives me around. That’s next level confidence in the system, especially in a place like San Francisco where the roads and people can be crazy. The seats don’t look to comfortable either.
The Race Heats Up
Recently the autonomy race has heated up significantly. Much like Tesla’s Model 3 accelerated the introduction of electric vehicles from all major automakers, Tesla’s Full Self Driving beta has sent a message to the nascent autonomy industry: You better get your ass moving. With Tesla putting their software into the hands of real customers, any player without a realistic launch strategy is starting to look like a loser.
You can see it on the streets of San Francisco. Autonomous test vehicles monitored by a human safety driver are becoming a much more common sight, with Waymo and Cruise ostensibly leading the way. Waymo has also started introducing some new vehicles. Previously they’ve always used Chrysler Pacifica minivans, but in recent weeks many people have reported seeing Jaguar I-PACE vehicles outfitted for Waymo as well. These fully electric vehicles are equipped Waymo’s latest fifth generation hardware.
So who will build a lucrative trillion dollar business, and who will die? Tune in the next episode of autonomy wars to find out. We’re getting close to the season finale now.