Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
AI Transportation

Hundreds Are Alreadying Using Waymo's Driver-less Taxis In Arizona (forbes.com) 115

The commercial rollout of Waymo's driver-less taxi service in Chandler, Arizona, a suburb of Phoenix with a population of 260,000 people, has more than a thousand customers already signed up -- including the mayor, reports Forbes: Each of the several hundred Waymo One vans in Chandler arrives with a safety driver at the wheel. But that may be more about public relations than technical necessity. During a recent trip, the human in the driver's seat didn't take her hands off her lap during a trip from the library to a shopping mall a few miles away in light, late morning traffic. "Part of it's just education and getting people really comfortable right out of the gate," a Waymo spokeswoman said. There's another piece of the Arizona program that's closer to Waymo's long-term plans of full autonomy. A few hundred people are getting rides in Pacificas with no safety driver through its Early Rider program, an earlier test rollout. Unlike Waymo One users, Early Riders have to sign nondisclosure agreements and aren't allowed to discuss the program.

Early Riders are also a way for the company to observe how people adapt to a robotic service and the options they want. Recently Waymo integrated Google Play music into the Waymo One app to let riders automatically listen to their preferred songs and artists. Video streaming, games and other in-vehicle options that leverage Google's many services are likely additions, though Waymo won't verify that... "Beyond the initial shock of not seeing a person in the vehicle, which we're getting used to, protocols are being established," says Chandler Police Chief Sean Duggan. "As a police officer, one of the first questions that gets asked is 'who gets the ticket? How do you contact whomever?'" There have been a "half a dozen" collisions involving a Waymo vehicle, Duggan says, but not ones where the Waymo vehicle was at fault. In fact, the department hasn't issued any citations to Waymo in the past couple of years...

Ahead of the commercial launch, there were reports that the vans irritate local commuters because they take too long to make left turns and of assaults on Waymo vans including rock throwing, a slashed tire and even an individual who aimed a gun at one. "People tend to be frustrated when a vehicle is actually obeying the law" by stopping completely at intersections and making turns cautiously, Dugan said. "That happens regardless of if it's self-driving or a person."

Forbes describes Waymo's presence in Chandler as "a test case for the entire industry," citing an interesting perspective from Bryan Reimer, a research scientist at the Massachusetts Institute of Technology's AgeLab. "The view for companies like Waymo is 'we have to be able to show functional safety. Otherwise, we can't protect our decisions in a court of law, where this will all end up long term.'"

"Elon is working mostly on the deep neural net side where a good chunk of it is a black box. Documenting, defending that in court is going to be tough."
This discussion has been archived. No new comments can be posted.

Hundreds Are Alreadying Using Waymo's Driver-less Taxis In Arizona

Comments Filter:
  • Fantastic (Score:5, Insightful)

    by JustAnotherOldGuy ( 4145623 ) on Sunday May 19, 2019 @06:49PM (#58620364) Journal

    I think this is fantastic.

    This will be a huge help for the disabled, people who can't get licenses (or who've lost them for whatever reason) the elderly, and anyone who's too tipsy to drive.

    For that matter, depending on affordability, I'd take one of these to work and back every day. I'd read or sleep or watch a movie on the way instead of dealing with traffic all the bad drivers out there.

    • Re:Fantastic (Score:4, Informative)

      by Rick Schumann ( 4662797 ) on Sunday May 19, 2019 @08:01PM (#58620624) Journal
      This will be a huge help for the disabled, people who can't get licenses (or who've lost them for whatever reason) the elderly, and anyone who's too tipsy to drive.
      So is calling a taxicab, Uber, or Lyft, driven by an actual human being.
      • So is calling a taxicab, Uber, or Lyft, driven by an actual human being.

        Sure, but sometimes those aren't available at some times and places. Ever tried to get an Uber at 2:45am Sunday morning in a residential area? Sometimes they just aren't there. Not everyone lives in a city where Uber drivers are circling the block like sharks 24/7.

        Also, some of the Uber/Lyft/taxi drivers I've dealt with are more than enough to make me want a driverless vehicle. And their driving often leaves a lot to be desired, frankly. I've had cab rides home from Seatac where I was pretty sure I was gonn

        • Why are you so opposed to the idea of driverless vehicles?

          Because new and scary, just like everything else techy. That's why China has nice things, and we don't.

        • Why do you believe the marketing department and media hype about so-called 'self driving cars' being so fucking good?
      • Re:Fantastic (Score:4, Insightful)

        by dgatwood ( 11270 ) on Sunday May 19, 2019 @08:34PM (#58620760) Homepage Journal

        The main difference is that it will initially be probably a third as expensive per mile, and is likely to eventually drop to around a twentieth the cost in the medium term. Also, RoboTaxi services are likely to have better response time, because apart from depreciation of equipment, there's not much cost to leaving a few extra cars scattered around the area, ready to pick someone up, unlike with human taxi services, where every extra driver is an extra mouth to feed.

        • by Anonymous Coward

          How can you justify the claim that driverless car services will cost 1/20th what they do now?

          We already have autonomous cars - they're called taxis. And the cost of paying the driver of those isn't 19/20th of the total costs of operating a taxi.

          • by dgatwood ( 11270 )

            A taxi costs three or four bucks per mile. An EV costs 15 cents per mile for power, plus depreciation on the vehicle. So if you compare a modern driverless EV to a taxi, 19/20ths of the taxi's cost comes from a combination of the driver, profits, and maintenance costs. Given that EVs are expected to have extremely low maintenance costs compared with gasoline vehicles, 1/20th isn't a completely crazy lower bound, particularly if you run the cars for three years and then sell them, kind of like the rental

            • An EV costs 15 cents per mile for power

              This is too high. Assuming a moderately-high power consumption rate of 300 Wh/mi, you'd have to be paying 50 cents per kWh for an EV to use 15 cents per mile in power alone. The average residential rate in the US is 12 cents per kWh, which would be 3.6 cents per mile. The average industrial rate is 6.7 cents per kWh, which would be 2.0 cents per mile. Given that operating an EV fleet would inherently provide lots of opportunities for time shifting -- and even more could be done by adding some stationary

              • by dgatwood ( 11270 )

                This is too high. Assuming a moderately-high power consumption rate of 300 Wh/mi, you'd have to be paying 50 cents per kWh for an EV to use 15 cents per mile in power alone.

                You're right. The paper I got that from was actually estimating 15 cents per mile in power + maintenance. My bad.

                That said, it is worth noting that a fleet vehicle will pay a lot more than 2 cents per mile for power. Bear in mind that 300 Wh/mile is only moderately high if you ignore the parasitic power use by the electronics. Those

                • Bear in mind that 300 Wh/mile is only moderately high if you ignore the parasitic power use by the electronics. Those parasitic losses are at their highest (up to 12 miles per day on Teslas) if you use software to frequently poll the vehicle for status information, send pickup requests, etc., which a RoboTaxi service would presumably do.

                  You need to do the math on that. Assuming the car drives 200 miles per day (low for a taxi), 12 miles per day constitutes a 6% increase. So 2.1 cents per mile rather than 2 cents. And as mileage increases it becomes even less significant.

                  If it drives ten miles per day, the parasitic losses effectively double the power consumption per mile.

                  If it drives ten miles per day, it's going to be completely uneconomical on depreciation alone.

                  And unlike consumer EVs, which can charge at night, a RoboTaxi service would need to operate all day, which likely means that they would need to charge during the day.

                  Peak costs aren't during the day, they're in the evening. Granted that self-driving taxis will do a lot of driving in the evening, but assuming they have 200+-mile batteries,

      • So is calling a taxicab, Uber, or Lyft, driven by an actual human being.

        Except for the cost of the driver, which makes these options unaffordable by most people for casual transportation.

        Driverless cars, obviously, have no driver. So they should be far more affordable once the market develops.

        Private cars spend 95% of their time parked. Driverless taxis can be actively driven far more than that, so the cost of the vehicle can be amortized over more miles much more quickly, lowering the likely cost of using them below even that of owning your own vehicle.

        For the elderly and di

    • Then a company, and shortly thereafter the government, will control where you get to go. Just think, the app can disallow any destination they choose. It will start with trivial restrictions: crack house addresses, known mafia hangouts, gun ranges, and similar noncontroversial spaces. Everyone will cheer as harm reduction is achieved. Then it will start getting more and more restrictive as people get barred from attending parties at the addresses of political dissidents, locations where protests are schedul
    • This is awesome. If this can become a real "thing", it will be a boon to the elderly, the visually impaired, and kids. You'd no longer have to rush out of work to get little Johnny or Susie to soccer practice. Nana can get to the doctor and do her grocery shopping without having to catch a ride. And best of all, now there will be reason for the braille buttons on the drive-through ATMs.
  • by Anonymous Coward

    the human in the driver's seat didn't take her hands off her lap during a trip from the library to a shopping mall a few miles away in light, late morning traffic.

    That's for all the trolls who will show up and claim this isn't "driverless". Yes, there is an employee on board to give the passengers some comfort that the company has "skin in the game" and is taking ample precautions. But the employee did not touch the controls.

    If you still think driverless cars aren't coming, you have gone into full ostrich mode.

    Here, let's go over it one more time.

    the human in the driver's seat didn't take her hands off her lap during a trip from the library to a shopping mall a few miles away in light, late morning traffic.

    • Re: (Score:1, Informative)

      BS. There was a "safety driver". That isn't driverless. And Tesla "Autopilot" isn't even autopilot. It is just a drivers assist, like what has been around for 20 years. Good luck with your VC hype. People are realizing that all this stuff is fake. That is why Tesla's stock is collapsing. No one is buying the story.

      • Comment removed based on user account deletion
        • I've already looked. It is a joke. Go take your hype and shove it. Tesla was bragging out a software update recognizing TRAFFIC LIGHTS last month. They still don't even recognize road signs. Volvos and Audis have had that for a decade. Ignorance. Tesla is dead.

          • Comment removed based on user account deletion
          • by Anonymous Coward

            Existence is pain. You're not my real mom.

      • BS. There was a "safety driver". That isn't driverless.
        Nope. And, it isn't because of 'public relations' or 'getting people used to it', it's because they don't even trust it themselves, and rightly so.
      • by galgon ( 675813 )
        Telsa's Autopilot is definitely more of a driver assist but it does more than others and it is better than others. This leads to people trusting it more, not paying attention and getting killed when they hit the side of a truck. Waymo's current tech is a beta driverless car. But they still have backup drivers so I cannot call it a self driving car. Does the car drive itself nearly all of the time - yes. Does it do it 100% of the time - No. We have California disengagement records for proof of this. Sa
    • Comment removed based on user account deletion
    • I have the feeling that "alreadying" is the latest abomination culled from gamer chat, and editordavid wants us to know he's down with the ten-year-old gamers. Which is ... a thing. I don't know if that thing belongs here, but what can you do except snark about it in the comments?

  • Why did a mention of Elon Musk get tacked onto the tail end of this story? Did Rei submit it?

  • "Each of the several hundred Waymo One vans in Chandler arrives with a safety driver at the wheel"
     
    You could have stopped there. Self-driving is a VC scam. And Tesla is going down. Sorry Rei. Bankwupt in 2 years.

    • Comment removed based on user account deletion
      • Total BS:

        "Each of the several hundred Waymo One vans in Chandler arrives with a safety driver at the wheel"

        Again, take your hype machine and shove it. No one is buying what you are selling any more.

        • Comment removed based on user account deletion
        • by Kjella ( 173770 )

          Different programs. Early Rider is their old invite-only NDA program, it's still ongoing and used to test features Waymo doesn't want people to publicly talk about. The only information we have about that is press releases and what can be observed by third parties, since anybody in the program would get sued if they talked about it. Waymo One is their first public service, initially it seemed extremely small and hard to get into but they've been expanding and as of now it always has a safety driver.

          They've

  • by Anonymous Coward

    But that may be more about public relations than technical necessity.

    To my knowledge, Waymo does 10,000 miles per 'interaction' on well-learned roads, but a human does 450,000 anywhere without getting in an accident. So unless Waymo has somehow multiplied their safety by a factor of 40, damn rights the safety driver is there for a reason.

    • Go stand at an intersection for a day and count how many times the bio-robots should be 'interacted' with. Just as with self driving cars, not every dodgy situation ends with an accident, so you cannot compare interactions with accident rates one to one, they are not the same thing.
  • by martyros ( 588782 ) on Monday May 20, 2019 @04:25AM (#58621720)

    "People tend to be frustrated when a vehicle is actually obeying the law" by stopping completely at intersections and making turns cautiously, Dugan said. "That happens regardless of if it's self-driving or a person."

    If people really get frustrated when a vehicle actually obeys the letter of a law for things like stop signs, that's a pretty clear indicator that the letter of the law is broken and needs to be revised.

    I've lived in the UK for over a decade now, and basically all non-traffic-light intersections are "yields"; which are effectively the same as a roll-through stop. It works just fine. The US rules around coming to a complete stop at a stop sign are unnecessarily strict.

    • by Guybrush_T ( 980074 ) on Monday May 20, 2019 @12:45PM (#58623884)
      The main problem here is that a self-driving car obeying the law might get stuck for hours at an intersection. There are times of the day when some left-turns will just never give you a 100% safe opportunity to pass. Those paths should never be used by self-driving cars but it's hard to predict. So I can totally see why people behind the car get frustrated. Obviously though, the rider will also get annoyed, so Waymo should just tweak their routing algorithm to avoid those left turns or improve their driving skills to better handle those hard cases.
  • I almost think that Slashdot being laden with typos gives it character, but I'm struggling to get past this mess of a headline. I'm sure I'll be corrected if 'alreadying' is actually a word. Thanks.
  • "People tend to be frustrated when a vehicle is actually obeying the law" by stopping completely at intersections and making turns cautiously, Dugan said. "That happens regardless of if it's self-driving or a person."

    They don't understand that there are two rules-of-the-road. There is the law, and there is the set of human rules and interactions that keep the traffic moving and are unspoken. When there are sufficient self-driving vehicles on the road everything will slow down, including business. The actual entire economy will theoretically slow down in the future. That can only negatively impact GDP and jobs. Think on that.

  • Any videos of autuonomous vehicles making left turns? A car that makes legal left turns instead of cuttig the turn short would be a positive step.

He has not acquired a fortune; the fortune has acquired him. -- Bion

Working...