Fully self-driving cars for hire have been hailed as the technology of the future for years now. That future has finally arrived — but only in some cities.
Transcript
ROB SCHMITZ, HOST:
Robotaxis, fully self-driving cars for hire, have been hailed as the technology of the future for years now. Has that future finally arrived? NPR’s car maven Camila Domonoske joins us right now. Camila, hey.
CAMILA DOMONOSKE, BYLINE: Hi, Rob.
SCHMITZ: So, you know, I keep hearing from my friends in San Francisco and Los Angeles that they are able to order these self-driving cabs, and they’re sort of, like, appearing everywhere. Are we going to start seeing more of these in the rest of the country anytime soon?
DOMONOSKE: Companies definitely have big plans for expansion. Currently, there are about a dozen big cities where these are everyday technology. A lot of them are Waymos. That’s really the industry leader here. These are these white electric vehicles with a bunch of sensors perched on top that kind of looks like a spinning top hat. You might have…
SCHMITZ: Right.
DOMONOSKE: …Seen them even in cities where they aren’t available yet. They’re doing testing.
SCHMITZ: Yeah, I saw one the other day in D.C. here, and it didn’t look like anyone is taking it for a ride.
DOMONOSKE: Not yet.
SCHMITZ: Yeah.
DOMONOSKE: But that is on the long list of about 20 cities that Waymo plans to expand into. I asked David Margines, the head of product management at Waymo, about this question of whether the robotaxi era has finally arrived.
DAVID MARGINES: I think the driverless future is absolutely here, and it’s rolling out to more and more places. It’s just not everywhere yet.
DOMONOSKE: And then there are other companies with plans, too. So Amazon’s Zoox just got approval to launch commercial operations. And then Tesla has big plans. They have designed a space-agey-looking Cybercab, but that’s not actually been deployed in the wild yet. In the locations – the couple cities where they’re running pilot programs for this planned robotaxi system, there’s more familiar-looking Tesla models that are doing the actual driving.
SCHMITZ: So how safe are these vehicles?
DOMONOSKE: Yeah, well, certainly the companies behind them have always said that they’re very safe. But in terms of third-party research, this summer, the Insurance Institute for Highway Safety looked at data from Waymo in order to track crashes per miles driven. And this is an independent group, but they were only looking at Waymo because Waymo is the only company that has publicly released the kind of data that they need for this analysis. David Kidd is the vice president for vehicle research at the group.
DAVID KIDD: What we found is that Waymo vehicles – they’re in crashes 68% less often than human drivers.
DOMONOSKE: So 68%…
SCHMITZ: Wow.
DOMONOSKE: …Fewer crashes overall. That is based on miles driven. And in terms of the kind of crashes, they’re also way less likely to be the vehicle hitting or striking another car.
SCHMITZ: That’s good to hear.
DOMONOSKE: Right. They might still get rear-ended by a human driver. Not so much you can do about that.
SCHMITZ: Right.
DOMONOSKE: Now, there’s an important caveat here, which is that this research was comparing Waymos to all human drivers, including ones behaving very badly. Many safety experts argue you should hold these systems to a higher standard than that. Maybe compare them to professional drivers or at least drivers on good behavior.
SCHMITZ: Got it. OK, so another concern that we’ve all heard is how many jobs could be lost among taxi and ride-hail drivers. You know, what other issues are people raising besides safety?
DOMONOSKE: There’s another category of concern, which is, these vehicles – they don’t make common human driving mistakes, like speeding or being distracted. But they can make kind of weirder mistakes.
SCHMITZ: OK.
DOMONOSKE: I was in Los Angeles earlier this year, and I called a Waymo. I wound up causing a real headache for some construction workers who were trying to get a big truck through an intersection. I was standing nearby waiting to get in the Waymo, and when it reached this intersection, it just stopped in the middle…
SCHMITZ: Oh, no.
DOMONOSKE: …Blocking everyone. One worker in a reflective vest was shouting and waving at it to try to get to move. That did not work.
(SOUNDBITE OF ARCHIVED RECORDING)
DOMONOSKE: This is my car. I’m so sorry. I’m the reason. I called it.
UNIDENTIFIED PERSON #1: Oh, wait, really?
DOMONOSKE: It’s waiting for me, but it says it needs to move to a new location, and it won’t let me get in.
UNIDENTIFIED PERSON #2: Watch yourself. Watch. Come this way.
UNIDENTIFIED PERSON #3: Welcome to the future.
DOMONOSKE: Welcome to the future. That was another worker there. It took about 4 minutes before it got out of the way.
SCHMITZ: (Laughter) Camila, that’s amazing that you had the presence of mind to take out your microphone for that (laughter).
DOMONOSKE: Radio reporter – always be rolling.
(LAUGHTER)
SCHMITZ: So, I mean, welcome to the future indeed. I mean, this seems kind of like a big problem – you know, when a machine has to manage a very spontaneous human problem in front of it. You know, so how does a company like Waymo improve its vehicles to handle issues like this?
DOMONOSKE: Certainly, more data helps over time is the argument that the companies say – that the more they drive, the more information they get to teach the vehicles what to do. David Margines from Waymo said, look, Waymos are not perfect and that they are actively working on improving how the vehicles respond in situations like this – which, I will note, I am far from the only person to have experienced, right? And it is particularly worrying when there are emergency response vehicles involved. That’s something that actually federal regulators have specifically raised the alarm about recently.
SCHMITZ: That is NPR’s very human driver and reporter Camila Domonoske. Thank you so much, Camila.
DOMONOSKE: Thank you, Rob.
Copyright © 2026 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.
Accuracy and availability of NPR transcripts may vary. Transcript text may be revised to correct errors or match updates to audio. Audio on npr.org may be edited after its original broadcast or publication. The authoritative record of NPR's programming is the audio record.
