back to article GM Cruise holds off on self-driving taxis for this year, says it needs more testing time to be safe

GM Cruise, the General Motors-owned autonomous vehicle startup based in San Francisco, won’t be rolling out self-driving robotaxis in 2019 after all, as the technology isn't quite yet safe enough. Dan Ammann, CEO of Cruise, said on Wednesday that the biz is still testing its vehicles on the city’s roads and will be ramping up …

  1. Disk0

    Still not flying either

    They should call the whole self-deluding car fantasy "Maybe After Tomorrow Or Sometime In A Far Far Future", or MATOSIAFFF for short.

    Instead just hang some fuzzy dice on your tv, use the pedalbin from the washroom to keep your foot busy and remember you are always commuting to the future in real time.

    Mine's the one that looks like a housecoat, because why go outside? There's lunatics there!

    1. revenant

      Re: MATOSIAFFF for short

      Nice, but BS is even shorter.

  2. trevorde

    Still waiting for...

    Self driving cars? Pffft! I'm still waiting for my flying car and personal jetpack I was promised.

    1. Mage Silver badge

      Re: Still waiting for...

      Why have trains still got drivers?

      It can't be JUST down to unions. Surely that can be done safely?

      1. murrby

        Re: Still waiting for...

        Iron ore trains in the Pilbara (where there are no hazards other than kangaroos and they won't bother a train that's 10km long) have just gone driverless at a cost of something like $500million. Tricky problem as they have to run at the right speed to load the carriages. On the other hand there's no stations and no passengers. And yes, it was done mainly due to unions (i.e. to cut costs)

        1. Ken Shabby

          Re: Still waiting for...

          When a driverless train which broke down on Sydney’s new metro line it had to be driven to the next station by a so-called “customer journey coordinator”.

      2. Francis Boyle Silver badge

        Re: Still waiting for...

        Well it's not that it can't be done because the Docklands Light Railway has been doing it since freaking 1987. More like the fact that saving the cost of a one driver per several hundred passengers just doesn't make up for cost of the infrastructure changes (which would be massive). Not that this has anything to do with the case for or against driverless cars. Trains are not driven by sight so the problems are quite different.

    2. John Brown (no body) Silver badge

      Re: Still waiting for...

      "Self driving cars? Pffft! I'm still waiting for my flying car and personal jetpack I was promised."

      You could try a jet powered hoverboard. Just don't use it over water....

  3. Anonymous Coward
    Anonymous Coward

    At least they have the right target

    "only deploy when we can demonstrate that we will have a net positive impact on safety on our roads"

  4. Doctor Syntax Silver badge

    "the technology isn't quite yet safe enough"

    As in the first 90% takes 90% of the effort and the remaining 10% takes the other 90%.

    1. Citizens untied

      This is reminiscent of a great Yogi Berra quote:

      "Baseball is 90% mental and the other half is physical"

      But he has another one that seems more relevant to this:

      "The future ain’t what it used to be."

  5. drand


    When you’re working on the large scale deployment of mission critical safety systems, the mindset of “move fast and break things” certainly doesn’t cut it.


    It's not just mission-critical, it's safety critical. Maybe that's just semantics but to me there's a big difference. There necessarily needs to be a lot of analysis and rigour in safety work but you will always come across issues in integrating sub-systems so there is a place for change it / break it testing of the products of your design; but that place is a controlled environment, not out on the streets.

  6. muddysteve

    Another 10 years away.

    I reckon we will get self-driving cars at the same time we get fusion nuclear reactors. It's always about ten years away.

    1. Jason Bloomberg Silver badge
      Thumb Up

      Re: Another 10 years away.

      I wouldn't put a timescale on either but it far exceeds the "we will have it done by next year" overly optimistic vision of many players.

      GM's attitude is positively refreshing.

    2. Chris G Silver badge

      Re: Another 10 years away.

      I suspect one of the factors delaying deployment is how to best monetise all the data they get from clients and how best to use that data while they are in the cab.

      " We are passing Happy Acres Cemetery, why not stop and make a reservation now? You never know what the future can bring!".

    3. vtcodger Silver badge

      Re: Another 10 years away.

      Pet peave: I've been hearing this "always ten (or 20) years away" stuff for decades. Sorry but I was around in the 1960s, and what I heard back then wasn't "ten years away" it was "60 to 80 years away". That'd be roughly 2020 to 2040. Right now it looks like 2050 or 2060 ... if at all. We might know more sometime around 2035-2040 when ITER is scheduled to be running actual full scale tests ... or not. I'm guessing that it'll take a couple of decades to translate success in the prototype -- ITER or whatever -- to actual commercial power plants feeding energy to power grids. And, BTW, while fusion power (if it happens) will probably be abundant, it may not be especially cheap.

      Autonomous vehicles are an entirely different story. It'd probably be possible to deploy a few today in VERY carefully selected venues. Shuttles of various sorts operating on dedicated pathways. The issue is how quickly that can be extended to more general environments where they must cope with pedestrians, human driven vehicles, bicycles, scooters, kids, dogs, (illegally) parked delivery vehicles, construction, accidents, cattle, demonstrations, weather, etc, etc, etc. I'm guessing that'll happen, but gradually and rather slowly. Decades, not months.

      1. Anonymous Coward
        Anonymous Coward

        Re: Another 10 years away.

        It's weather that's the killer (so to speak) because bad weather really messes up LIDAR. Strange that all these self-driving car trials are in California, isn't it.

      2. Mage Silver badge

        Re: Shuttles of various sorts operating on dedicated pathways

        Those are called trains.

        Some trams/trollies might qualify, but while hugely simpler than a car (much more defined route) there are still huge safety issues. Who's going to flick the switch to save five and kill one? Of course if the face recognition identifies the one potential victim as a serial <insert action here>.

  7. tiggity Silver badge

    Not in my lifetime

    Would be my bet

    Would be useful as you get old and reflexes / brain in general slows down so you're too unsafe to drive* but AI cars could get you around (vital if live in an area with little public transport)

    But I'm expecting a dotage of expensive human driver taxi journeys instead

    *GPs seem distinctly unenthusiastic to contact DVLC to get peoples licences revoked - seen elderly relatives with degenerative issues (e.g. Parkinson's) allowed to drive far longer than it was safe (in one case police, not GP, finally caused licence revoke as he was stopped by police for unsafe driving)

    1. batfink Silver badge

      Re: Not in my lifetime

      That's no way to talk about the Duke of Edinburgh....

  8. 89724102172714182892114I7551670349743096734346773478647892349863592355648544996312855148587659264921 Bronze badge

    Self driving trains would have been far simpler to start with

    1. LoPath

      They already have some around here that are remote controlled running around train yards. The dream job of every kid that loved playing around with toy trains!

  9. Charlie van Becelaere

    Quelle surprise

    Quelle surprise indeed.

    Still, I'm very pleased with their decision.

  10. c1ue

    I have 2 in-person stories of Cruise cars on the road in SF.

    The first: the San Francisco Muni bus I was riding on almost ran over a Cruise car at an intersection. The intersection is a complicated one - there are 5 passages out of it, none of which are 90 degrees to any other. We all started at a red light. The Cruise car in front was apparently choosing to turn to the 2nd left, with the bus I was riding on behind it. The Cruise car turned - all seemed fine until it slammed on its brakes about 2/3rds of the way through the intersection. There were no other cars in front or beside it. There was no obstruction. The bus wasn't moving that fast, but wasn't moving that slow either (it is an electrical trolley bus).

    The second: I had rented a car to go to a meeting 50 miles away. Since I had the car, I went to the grocery store since I had time before the meeting. When exiting the grocery store parking lot - which is a pain because cars parked along the road, due to the slope, block vision for incoming traffic coming from the left.

    I pulled forward a bit, nearly to the roadway, in order to see. A Cruise car is coming along. It turns on its right turn signal like it is about to enter the grocery store parking lot. I wait, because human drivers are not trustworthy - fortunately because the Cruise car not only didn't turn, it accelerated and edged closer to its right to almost hit me.

    Note both of these aren't even the normal heavy traffic, pedestrians staring at cell phone, dogs and homeless people meandering across the street, giant pothole, flying trash, Uber/Lyft car stopped anywhere types of city SF driving situations.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2020