Recently at the CES show in Las Vegas, I got a chance to ride in the Zoox Robotaxi.
Zoox doesn’t get a lot of talking about compared to Waymo, which is the clear leader, and even Tesla, which doesn’t even have a working Robotaxi, but of course, is a very famous company. I came to meet Zoox actually before it was founded, when the founder, Tim Kentley-Klay, came to me to discuss robotaxis and robocars and his plans for them. I decided not to join them, but had various discussions and debates with him.
He wanted to put forward a very different philosophy of a robotaxi startup. That involved building a vehicle that was symmetrical in all four directions, front to back and left to right, custom-built from the ground up to be a robocar–not like most other teams were doing, which is taking an existing car, modifying it, adding sensors and computers, and trying to turn it into a robocar or robotaxi.
He felt that it was time, even then, for “what comes beyond the car.” Some vehicle which is designed specifically to be that. My Zoox ride in Las Vegas was unfortunately interesting, which is to say it was not perfect, which is not a good sign for Zoox. After all, when you want to make a robotaxi with no one in it, you’ve pretty much got to do a perfect job on a randomly sampled ride. Robotaxis all make mistakes and will continue to make mistakes for their whole lives. But those mistakes must be very rare. If you see one on your first ride, it’s often a very bad sign.
You can’t in fact, even tell if a robocar is any good by riding in it yourself for one ride, two rides, 10 rides. Even 1,000 rides isn’t enough to know that it’s good. Unfortunately, one ride can be enough to know that it’s bad. Zoox is still in an early stage, and that’s very clear from the service that they’ve started in Las Vegas, which only runs between about eight different locations and over an even smaller number of routes between those locations. You can only pick these things in the Zoox app. In San Francisco, they have a more anywhere-to-anywhere application, which suggests they do have some of this maturity.
The Zoox vehicle, as noted, is custom-designed. It’s fairly sleek. It’s comfortable. It’s clean. The lines are nice. I did find, as many others have found, that the roof is a bit low when you’re getting in, particularly if you’re getting a little creaky in your bones. And that may or may not interfere with some people’s enjoyment of the vehicle.
Play Puzzles & Games on Forbes
One thing that’s quite striking is that this symmetrical view basically doesn’t have a windshield for viewing out the front or the back. There is no front or back. They’re both the same. That means you can only see what’s in front of you through a very small window. You do actually have a very nice view out the sides, where the doors are quite tall. It’s sort of what you might expect for a tourist vehicle or something like a train, where you of course, can’t see out the front. You only see to the side.
The seats face each other, so the Zoox is much more social when you have a larger group, or even a group of two, than any of the other vehicles if one person sits facing forward and the other sits facing back. Not everyone likes facing backwards, so that could be an issue. However, with four seats, you’d need three people who can’t stand facing backwards before it was ever trouble. That’s actually not that likely to happen because only a small percentage of people get a little motion sickness from this. I didn’t have any problems on my backwards ride in the Zoox. I talked to some people who said that they often are bothered riding backwards, but not in the Zoox, so maybe they’ve done something a bit better when they put it this way.
It’s a little small if you’re going to have luggage inside this vehicle. There’s no trunk. There’s just the passenger compartment with the two people sitting opposite each other. Some of the features included in the Zoox to make it different including the fact that since it has no front or back, it can change direction without really doing anything. It just has to reverse. It’s even equipped with four-wheel steering, which means it can do some tighter maneuvers, move into parking spaces better than a classic car can.
Unfortunately, on my ride, as I said, we encountered some problems. The most distinct was a strange incident where the vehicle did what’s called phantom braking, which is stopping when there’s nothing there, or sometimes called a false positive. The vehicle was turning a corner onto the strip, Las Vegas Boulevard, and it hit the brakes hard and then started moving again, hit the brakes hard again immediately, and did that three or four times. Zook said it was worried about a school bus, which I didn’t see and couldn’t have been very close to us. Phantom braking is a problem that all vehicles have because you have to decide whether you’d rather brake for nothing or not brake for something. That decision is very easy. This means that you make the vehicle a little more sensitive than maybe is ideal, which means it will phantom brake.
That’s okay, especially when you’re starting out, because you don’t want to hit something. It’s okay if you just disturb people every so often. It’s not okay if you phantom brake and a car behind you rear-ends you. That’s legally the fault of the car that rear-ends you, but it’s still not a good thing to make happen. This is definitely a sign that Zoox has got some work to do. Note that we do see Waymos and Teslas and other vehicles making phantom stops.
I was more worried for a while about another issue, where, for a time at least, to me it seemed as though the car had run a red light. The light turned red quite some time before we went through it. Examination of the rather strange intersection where this happened, where the light is mainly for pedestrians, showed that the gap between the stop line and the red light itself is actually quite long. And this meant that we were past the stop line well before the light turned red, which is the law in Nevada and most places. It was still a little bit disconcerting.
Will these features of the Zoox make a difference in the race? Zoox has put a lot of effort into making its vehicle have this unconventional design. It’s put a lot of money into it, and it’s paid a very high price for it. Is symmetry that important? Zoox says that, for example, the vehicle could pull over to pick up passengers, and if it’s going back the way it came, it doesn’t have to turn around. It can just go back the opposite way.
(The Zoox doesn’t show you any way, which is going to be the front and which is going to be the back when you get in. If you do care about whether you ride forwards or backwards, and most people do prefer to ride forwards when they’re alone or with just two people, you won’t know where to sit.) You would have to switch, which technically you can’t do because you have to be seatbelted to ride in the Zoox. They might put in some user interface for that in the future.
This symmetry does allow a few maneuvers for the Zoox that other cars wouldn’t do as well. But any car, any robocar at least, can drive backwards just fine. Robots are perfectly capable of driving backwards even when the wheels that are steering you are now in the back, if the lights are configured for this. It looks very weird when a regular car does that. The Zoox doesn’t look weird when it does this. But that would be something that might bother other drivers for the one or two blocks that you would go backwards before you turn yourself around by pulling into a driveway or some other place. As such, I’m not sure there’s going to be that big an advantage if it becomes something really important in competition.
Zoox is hoping that people will pick their vehicle over others when it’s time to do so because it has these special features. It’s less clear they will, at least not for many years. The time will come when we’ll see actual head-to-head competition between the robotaxis. They’ll be fighting each other on features and price. But that day is not today, and it’s not for several years. Zoox actually may have done this a bit too early.
You pay a big price to design your own custom vehicle. Designing a new vehicle is really hard. It’s a lot of work and a lot of money. And Zoox put tremendous effort into making this vehicle. And the effort shows–but what’s it for, at least today? Down the road, maybe it will be for something. Kently-Klay was very keen on being redundant with the symmetry. There are four motors and everything in the vehicle is duplicated, and so it can handle individual component failures.
That is useful, and all the other robocars also include some level of redundancy. It turns out mechanical failure isn’t a big problem in a robotaxi compared to a private car. When a private car breaks down, that’s very annoying. You’ve got to get a tow truck, you’ve got to get to a dealer, they’ve got to get you a loaner car. Your whole life is disrupted. Thus, you pay a lot so that the private car you buy doesn’t break down, and you’re very angry at the company that sold it to you if it does. But if a robotaxi breaks down, not in a safety way, but just in a mechanical way, they can send another robotaxi and have you on your way in just a few minutes. The total inconvenience is quite small, and you’ll probably get a free ride coupon when it happens, so that it’s actually a win for you, perhaps.
Did Zoox chase the right things in making their vehicle so expensive? The extra work did result in some great problems for Zoox. They didn’t get to develop as quickly as they wanted to. They ran out of money. They couldn’t raise more money, so they ended up selling themselves to Amazon.com. Amazon is a great parent. It’s got lots of resources and money. It is one of the few companies that has the resources to put a robotaxi into the marketplace. Through Amazon Prime, it has hundreds of millions of people who’ve already got accounts with Amazon who can easily be marketed into using the vehicle.
That’s all great, but for the founders, it meant having to sell much earlier than they planned to sell, which I don’t think they wanted. There’s even some speculation that when that founder, Tim Kentley-Klay, was forced out of the company, it was in part because of his dedication to this very expensive mission and the management style that went with it.
Zoox is on the roads today, which is a big achievement. Only a very small number of companies have actually put a robotaxi on the road with no one in it. That’s something even Tesla has only pretended to do. Zoox gets big props for that, even though its service today is quite limited in scope. With the money that Amazon has, it can probably grow it to be bigger.
It’s good that we’ll see competition in the robocar space. We don’t want to have just one company today, like Waymo, in charge of everything. Zoox is not in production yet. We can hope that they will be soon.
A few years down the road, their design efforts may pay off. It may give them an edge. The other companies, if they see that turns out to be true, if they see customers are switching over to Zoox because their vehicle is just a little better at the job, because it’s designed specifically for the job, others can adapt new vehicles too.
We know that Cruise, when it was alive, with General Motors, was building a vehicle that actually looked a lot like the Zoox in many ways, though it wasn’t fully symmetrical. The Cruise Origin had a roughly similar shape and the face-to-face seating, and many of those other features. They intended a custom vehicle. Tesla, while it doesn’t yet have a robotaxi, is already finishing the design of a vehicle called the CyberCab, which has only two seats in order to be inexpensive, and will be a custom-designed robotaxi. That’s going to come, and Zoox is going to face that competition. Nonetheless, we should continue to track them and hope for their success and the success of all teams in building the RoboCar vision and robotaxis and giving them to the world.
