This week, Cruise Automation, the self-driving car subsidiary of GM, introduced Origin, a fully autonomous vehicle that has no driving controls whatsoever. It’s meant to be a rolling pod that carries passengers on demand, almost like a small bus or train car. 

Origin is on the way. But are companies allowed to operate cars without steering wheels on public roads? Both Cruise and Waymo have pushed the federal government to lift requirements on equipment — like pedals, steering wheels and mirrors — and these vehicles are allowed, in certain conditions. States have their own rules. Although carmakers and safety advocates have been hoping for some clear guidance on what will and won’t be allowed nationwide, Jack Stewart, who covers transportation for Marketplace, says that’s not coming anytime soon. 

We broke it down in this week’s “Quality Assurance,” the segment where we take a deeper look at a big tech story. The following is an edited transcript of our conversation.

Jack Stewart: The federal government is indicating really strongly that it is just taking a hands-off approach to regulation here. In fact, Elaine Chao, the transportation secretary, just spoke about this at CES and said that this was going to be left to the companies developing these vehicles to self-regulate. There are two views on that, depending on how you look at regulation, I suppose. There are things that hold back development, and laws and rules that are set in place in a very fast-moving environment can become impediments to progress. But also, there’s an argument that safety will not be as well scrutinized if it’s left in the hands of the companies — and certainly some people who are concerned about the safety of these vehicles are saying that we need stronger regulation.

Molly Wood: We should be explicit here, though: The National Transportation Safety Board, auto safety advocates, have strongly said there needs to be federal regulations for testing these vehicles and that it shouldn’t be voluntary.

Stewart: The NTSB came out, in particular, after we saw the incident with the Uber self-driving vehicle kill a pedestrian pushing a bike. It’s accidents like that that could set this whole industry back. If we do see more fatal accidents, then I think we’re going to see a really big hit in public confidence in these vehicles. That is something that federal regulation could help with, I think. It’s really just at least building confidence in the safety of these vehicles, and is there somebody, somewhere without a vested interest in them who is scrutinizing what is going on here?

Wood: Back to Cruise. When they design something like the Origin, which doesn’t have any of those controls, do you think that that’s a bet that the light-touch regulation is going to continue?

Stewart: I think the Cruise vehicle might fit into a slightly different category. We’re seeing a splintering in the way that these autonomous vehicles are being developed now. Waymo is going all in on running a complete taxi service. It’s a type of Uber service. They’re already up and running in a very limited form outside Phoenix, Arizona. This Cruise vehicle looks to me like it could operate in a more constrained environment, which makes the problem just a little bit easier to solve. If it can operate at lower speeds in cities and that type of thing, it’s a slightly easier problem to crack. There’s different parts of this autonomous vehicle problem that might be easier to crack than just going all in on this, “We will make a car that can drive itself in every single circumstance.”

Wood: Right now, it seems to be mostly a concept. Did Cruise announce a service that goes along with this?

Stewart: Cruise is stressing that this is a production-ready vehicle, but without giving any details about how it would go into production. There would be a large number of hurdles to overcome, not least the fact that it doesn’t have human controls, which means they can’t go on the roads as a car at the moment. Certainly not in the U.S. There are still so many rules about having to have a steering wheel; having to have a brake pedal; having to have mirrors on the outside. All things that are very useful for humans, but pretty useless when a computer is in charge. The whole thing is decked out in sensors, and you don’t really need any of those controls. That’s not going to happen anytime soon, but Cruise is already doing a lot of trials in the San Francisco area with much more conventional vehicles — Chevrolet Bolts, the little electric cars, which it has equipped with its sensor technology, and it is running trials around the city and releasing its various updates on how it’s doing. It seems to be progressing quite fast. It’s up there with the Waymos of the world. It’s not quite running a commercial service yet, but nobody’s really quite delivered on those promises yet either. It’s one of the companies to watch for sure.

Wood: Look, no new technology exists without pushing the envelope. But are people ready or even close to ready for cars that do not allow for any human intervention?

Stewart: I guess we won’t know the answer until people actually start riding in them. What I’ve seen of the types of autonomous vehicles that are on the roads now is people really quickly get used to it and almost become blase about it. You look at the videos that Waymo releases of people riding in their cars. The first time they get into one of these vehicles, it’s like, “Wow, this is amazing, the steering wheel’s moving by itself.” Then, like the third time they get into one of these vehicles, they’re just looking at their phone and letting the computer deal with the traffic and not really worrying about it. I think as long as these companies can maintain a decent safety record, then people do have the tendency to become pretty used to things pretty quickly.

Wood: But you still have this action movie idea that if you needed to, you could leap over the seat and start driving.

Stewart: I’m guessing that these vehicles will still have some giant, red button in them that you can press to halt everything. In fact, to test in California, eventually any vehicle that does not have a driver will have to have various safety protocols, including ways to contact maybe a central control center or something like that, so that somebody else can dial into these cars. They’ll have to have ways of communicating with law enforcement officers, because remember, if there’s no driver and this thing gets pulled over, who’s the cop at the window going to actually talk to? There’s a lot of things that will go into place to help people feel more comfortable in these cars and hopefully keep them a little safer.

Wood: Mercedes had a concept car a couple of years ago — it did have a steering wheel and a driver’s seat, but it was this pod where the seats all faced each other. I have to tell you, I rode in that thing, and I almost barfed. This is a serious issue. Is anyone talking about the car sickness factor here?

Stewart: I haven’t heard a lot of discussion about that yet. I don’t know. Was there something about it that made it worse than just having a bad human driver? Was the car driving badly, or was it the way you were facing?

Wood: You’re not necessarily facing forward in the same way. It didn’t have windows, it just had screens to show you things to create an atmosphere. It’s a slightly different driving style, because you really are being driven and facing the wrong way. Maybe people feel that way in limos, too. Virtual reality faced this hurdle and didn’t see it coming, so I’m just trying to pinpoint all the possible fail points.

Stewart: People really want to be able to get rid of the windows and project onto those. There are a number of companies that are already planning to show us movies and adverts and all sorts of things in these autonomous cars once you don’t have to focus on the road. There’s going to have to be something that they deal with. Facing backwards is another issue. There’s no reason to face forwards, necessarily, if you’re not looking where you’re going, but some people really hate that. Some people are fine with it. Sometimes on trains and planes, people enjoy facing backwards, and it’s actually slightly safer if you’re in a collision to face backwards to that collision. I think it’s just going to have to be part of the adaptation that we all go through as these vehicles become more common.

Wood: Barf.

Related links: More insight from Molly Wood

One interesting note is that it’s not just safety advocates who would like some clear regulations. Carmakers feel like a nationwide standard would benefit them, too, since some states have much more stringent rules that could prevent them from rolling out, for example, a taxi-like service that depends on little rolling pods with no steering wheels or pedals. 

I’m not letting go of this motion sickness thing. Remember, that was something virtual reality did not completely anticipate, and it was a real problem. Fun fact: It turns out that women are much more likely to get motion sick than men — perhaps one of the reasons why it was not anticipated by the heavily male-dominated VR industry. 

A story in The Atlantic talks about how we don’t totally understand why some people get motion sick and others don’t. The story notes that “certain groups are more likely to get queasy,” and those certain groups are kids and women. Basically everyone who’s not an adult male? That seems like a pretty big group. 

I’m just saying that if all these people are out here creating airless pods where people have to sit backward, and they want to get rid of all the windows and replace them with holograms and probably advertising and scenes of a nicer outdoors than we actually have, and every woman or kid who gets in one of those cars literally boots, it’s gonna be a problem.
Luckily, I am not the only person who’s thought of this. There is one major study in America from the University of Michigan looking into this issue, based on the idea that it could in fact hinder adoption of autonomous passenger vehicles. The team is run by Monica Jones, who’s gotten carsick all her life. They published that research last fall and created the first large-scale methodology for studying driving maneuvers and activities that might make people sick. Heart you, Monica. You see me.

Stories You Might Like

How autonomous carmakers can make their cars more accessible

05/02/2018: Self-driving cars still need our help and that might be a problem

Our driverless car future

Apple seeks more control of vehicle software with CarPlay’s new features

There’s no driving test for self-driving cars — only metrics

The autonomous vehicle industry hits another roadblock