Catch up on stories from the past week (and beyond) at the Slashdot story archive

 


Forgot your password?
Close
typodupeerror
×
Japan Transportation

Nissan Leaf Prototype Becomes First Autonomous Car On Japanese Highways 140

Posted by samzenpus from the keep-you-eyes-on-the-road-your-hands-off-the-wheel dept.
cartechboy writes "As car manufacturers battle over futuristic announcements of when autonomous cars will (allegedly) be sold, they are also starting to more seriously put self-driving technology to the test. Earlier this week several Japanese dignitaries drove — make that rode along — as an autonomous Nissan Leaf prototype completed its first public highway test near Tokyo. The Nissan Leaf electric car successfully negotiated a section of the Sagami Expressway southwest of Tokyo, with a local Governor and Nissan Vice Chairman Toshiyuki Shiga onboard. The test drive reached speeds of 50 mph and took place entirely automatically, though it was carried out with the cooperation of local authorities, who no doubt cleared traffic to make the test a little easier. Nissan has already stated its intent to offer a fully autonomous car for sale by 2020."
This discussion has been archived. No new comments can be posted.

Nissan Leaf Prototype Becomes First Autonomous Car On Japanese Highways More

Nissan Leaf Prototype Becomes First Autonomous Car On Japanese Highways

Comments Filter:

  • Japansportation.

  • > who no doubt cleared traffic to make the test a little easier

    There are lots of empty roads NorthEast of Tokyo, and not having a human in the car is actually recommended.

  • I think people just won't own these cars (Score:3)

    by GoodNewsJimDotCom ( 2244874 ) on Wednesday December 04, 2013 @08:06PM (#45603741)
    I think for a person to own a self driving car might be the exception to the norm. I think if self driving cars work, corporations will buy millions of them, and station them in semi-patroling routes. Then people will just summon them like cheap taxis. Some people will even schedule their work day around them. The software will do all the planning on who gets what car. A guy could ride one to work, not pay parking, then the car plays taxi for the day, and comes picks the guy up at work to take him home.

    If they work, they'll work big time, but I really worry about lawsuits.

    • Meh, there are just too many people for 2D movement.

      Someone had better figure out how to reliably send me a drone to taxi me around.

      • The subways in Seoul, Moscow, Beijing, Shanghai, and Tokyo together provide over 12 billion rides per year (cite [wikipedia.org]). Tunneling is totally the way to go for dense transport. It's 3d, safe, efficient, relieves the open land of the blight of travel infrastructure, and the high initial cost of tunneling is super-amortized when ridden billions of times per year.
    • One of the biggest potential long-term issues with these cars will be maintenance. If you look to autopilot in airliners as an example, autopilots only tend to fail/disengage when there is something wrong with the sensors on the plane. For instance if the pilot and co-pilots air speeds disagree(indicating a potential blockage in a pitot tube) etc. Now since the airline industry is heavily regulated and since airplanes are so expensive, the airline companies have a huge incentive to keep their sensors wel
      • I don't see that as any different than someone not properly maintaining their manual car. I know lots of people that have been in accidents because of bad breaks or bawled tires.

        • Re: I think people just won't own these cars (Score:5, Funny)

          by Gravis Zero ( 934156 ) on Wednesday December 04, 2013 @09:13PM (#45604143)

          I don't see that as any different than someone not properly maintaining their manual car. I know lots of people that have been in accidents because of bad breaks or bawled tires.

          people shouldn't be taking a break while driving much less bad ones! how can they even stand with crying tires? dude, you know some messed up people.

          • I'm sorry for that, I was writing on my phone and /. mobile really sucks. I have a Galaxy Note II, writing with a stylist, phone does auto correct and when I go back to correct spelling the phone sometimes goes into epileptic seizure with the scribe area popping up and disappearing randomly, conveniently over the submit button. There's no preview on /. mobile, it just posts.

            • Re: (Score:2)

              by JazzLad ( 935151 )

              writing with a stylist

              There's your problem. ;)

              • I guess I should have made sure she actually graduated high school before letting her cut my hair and write my posts. That was all my mistake, sorry.

                • Re: (Score:2)

                  by JazzLad ( 935151 )
                  Kidding aside, you did make some excellent points elsewhere in this thread - specifically regarding people doing makeup (or texting, whatever) being far more dangerous than automated cars.
                  • Thanks. I do hate to sound militant about it, but I've had a lot of close calls driving and 99% of the time it's because of other people and it would have been entirely preventable if those people were to just do what they're suppose to be doing. So I just can't see how this technology could *possibly* be worse than the existing situation.

                    Who hasn't been driving down the highway and seen someone texting, or shaving, or putting on makeup, or screaming at their four year old in the back seat. We just have t

        • Except now the car can just take itself to a maintenance appointment while you are at work or overnight, so you never need to actively actually do anything. In any case, I think the cars will make the best estimate of the world that they can, based on a combination of sensors. It seems to be the case that you can drive a car optically (humans do it) so if the radar sensors go out, it's probably still perfectly safe to let it drive on lidar and cameras for a little while. It doesn't get scary until cheap e

        • I think the main difference is maintenance cycles, delicacy of sensor and detection of failure.

          Brakes don't suddenly go from good to bad. They have a very graduate wear and it's easy to detect that they should be replaced in the annual checkup. And when it is detected, there's plenty of time usually left, not to mention that you notice a change in the behaviour unless you're very insensitive to your car's signals.

          Likewise, if you're lacking oil, it's trivial to detect that. There's a sensor that notices whe

          • Brakes don't suddenly go from good to bad. They have a very graduate wear and it's easy to detect that they should be replaced in the annual checkup. And when it is detected, there's plenty of time usually left, not to mention that you notice a change in the behaviour unless you're very insensitive to your car's signals.

            Is it just here in NZ that brakes are typically of the "screamer" variety? I can't imagine we've any monopoly on them here. I occasionally hear them in cars around the place, yelling like tortured pigs because the meat on the brake pad is low. Very embarrassing and a good incentive for the owner to get them changed pronto. Happily, they don't seem to make any difference to the braking performance. It's quite a different noise from that one running the brake pads down to the metal and it certainly is rather

            • We do have those in Canada. I primarily hear them on Metro Transit buses around my city, which is kind of unsettling.

          • The brakes on my car went from good to bad rather quickly when a seal failed in the slave cylinder, likewise my power steering failed entirely when a belt to the pump snapped.

          • Brakes don't suddenly go from good to bad.

            Tell Paul Walker.

            In fact, brake lines fail without notable warning all the time, as do other components like masters and boosters. So you're wrong there.

            Likewise, if you're lacking oil, it's trivial to detect that. There's a sensor that notices when there isn't enough oil and it works trivially easy. Covered in oil = fine, not covered in oil = warning light on.

            Actually, this isn't that easy. For example, one of the UPS drivers let me know that his Mercedes Turbo-Diesel powered delivery van was detecting low oil and shutting off when he needed power the most, going up bumpy hills. Almost killed him one time. The fleet mechanic defeated it for him so that he wouldn't die. So you're wrong again. Even Mercedes who ha

          • I live in Atlantic Canada and have had brakes seize because of salt corrosion on the calipers, which isn't a fun thing to discover when you live in a city that's primarily one big hill. I already had a appointment to have my brakes serviced for the following day, but I wish I hadn't put it off that long. Luckily no one was hurt when I ran a stop sign and was able to eventually stop the car using the emergency brake and pushing really hard on the pedal. The brakes engaged, but then seized in the engaged posi

          • Re: (Score:2)

            by Kjella ( 173770 )

            Brakes don't suddenly go from good to bad. They have a very graduate wear and it's easy to detect that they should be replaced in the annual checkup.

            The brake pads themselves, generally no. But you can have catastrophic failure in the brake fluid tubes and with no pressure, next to no breaking. Still overall mechanical failure is the reason for very few accidents compared to human error.

        • I don't see that as any different than someone not properly maintaining their manual car.

          I see a big difference: with an automated car, the car will know that it needs maintenance. If it is a safely issue, it can limit its speed, or refuse to drive until the problem is fixed. Otherwise, it can automatically drive itself to a maintenance center while you are at work or sleeping.

          • Like when you have a stroke. Your brain knows that it is impaired and you go get medical attention.

            How many layers of sensors and redundancy will be enough?

            • The computer the sensors are connected to will know when the sensor isn't functioning. Cars from the last seven years, at least, have had this feature along with just about any machinery on assembly lines that work at high speeds. When the brake sensor on my 2007 Yaris goes a light tells me I need to take my car in.

              Aside with that what's wrong with redundancy? What you don't carry a spare tire in case you have a blowout while you're on the highway? Seems like a good safe idea to me to make sure there's an

          • Re: (Score:3)

            by AmiMoJo ( 196126 ) *

            It's gonna be weird seeing cars and trucks driving around with no-one in them at first. Could open up some new and interesting avenues in highway robbery too, especially if all you have to do is set up a stop sign in the road and the computer automatically brings the vehicle to a halt.

            • Could open up some new and interesting avenues in highway robbery too

              Doors are locked. If a window is broken and a door is opened, the car drives to the police station.

      • Not a problem...lots of biz models
        1. insurance includes mandatory and included in your premium sensor/systems maintenance
        2. car is subscribed to or leased which includes maintenance in monthly fee
        3. cars are not owned; just used like taxis; so they're maintained by a company under strict regulations ...and the cars will maintain themselves. a loaner car can drive over while your car is worked on.

        • 1. insurance includes mandatory and included in your premium sensor/systems maintenance

          The advantage to this would be everyone pay's the same rate, it won't matter if you're male or female, young or old. All cars will have the same, tiny, equal chance of failure so the insurance company won't be able to get away with the, "You're a male so you pay twice as much as your wife, even though she's had three claims in the last year and you've never had so much as a warning in your life."

    • Re:I think people just won't own these cars (Score:5, Insightful)

      by timholman ( 71886 ) on Wednesday December 04, 2013 @09:07PM (#45604119)

      If they work, they'll work big time, but I really worry about lawsuits.

      I tend to think the lawsuit fears are overblown. In the U.S. alone, 35,000 people die each year due to human drivers, at a cost of about $200 billion annually, paid for by everyone's insurance. We seem to have no problem living with that.

      If autonomous cars can cut that fatality rate to 3,500 or even 350 deaths a year, the savings will be so enormous that it will be cost-effective for the auto companies to partner with insurance companies and create a general fund to reimburse those people who may be injured due to an automation failure, regardless of fault. The federal government already uses this concept with the National Vaccine Injury Compensation Program. It provides no-fault reimbursement of vaccine-related injuries, because letting vaccine makers be sued out of business would result in more deaths and injures in the long run.

      And keep in mind that accident rates will only continue to drop as the automation improves with time. Moore's Law is inexorable.

      • You know what makes me think it will fly:
        Since big time corporations will invest in them, they'll have big time corporation pockets to win lawsuits.

        I think relative safety is important in the end, but you know how court cases are, logic doesn't always prevail.

      • what about legal liability as tickets up to criminal stuff that can get you locked up.

        What if a auto car hits someone? and drivers away as some sensors failing may read it as some thing that is on the safe to drive over list.

        • This stuff can happen today, eg. brake failure, so it's not unprecedented. In the case of brake failure, the liability is typically in the hands of the manufacturer and/or the dealership. Only extremely rarely is that a jailable offense (in the sense that vaccination injuries would only very rarely be a jailable offense).

          This isn't a new kind of liability problem, it's just a different scope.

      • very well written, and I agree completely.

        The question is, will they put all those cash savings towards the loss in revenue from speed traps and red light cameras?

        Me? I'm betting they find new ways to fine us.

      • I think he was more worried about what would happen to all the poor lawyers when there is a massive drop in lawsuits. Somebody think of the lawyers!

      • shared insurance liability could be a great way of getting over what are likely to be a small number of high profile law suits in early days of adoption. as you say works great for vacination.

        on side note people always quote the deaths from auto accidents. don't forget to multiply the figure by about 3 for accidents resulting in serious injury where long term care and support needed.

    • Just wait for cab driver unions to go berserk over it. It kills jobs! It's hackneys and RIAA all over again. Just you wait and see, these things will be outlawed inside the cities.

      • Sorry to reply to you twice in the same thread but it seems to me that this is going to be a common refrain over the coming years. Taxi and transportation companies ought to be thinking hard about the future by now, knowing the impact of autonomous vehicles will be gigantic. Especially so when accounting for the extra pressure the potential of drone deliveries brings. That's two strong areas of development (that have already proven themselves acceptably reliable and safe even this early on in their lifecycl

    • Some people will even schedule their work day around them.

      That would be counter to the whole point, imo. They'll be taxis, not busses. And with smartphones, easier to summon than ever. Although I agree ownership will decrease in urban and semiurban areas.

    • Sounds like a bus to me :)
      • Except the car won't leave until you're ready, you don't have to share it with 50 other gross diseased people hacking and coughing all over you, you'll actually be able to take your coffee to work with you, you won't have to make 100 stops between where you work and your house, the vehicle will go from point A to point B without having to drive all over the city first, you can get dropped off at your front door rather than a stop two blocks away, the last three mean the trip will be 15 minutes instead of 2
  • ...with live streaming to web!

  • Automotive networks have zero security (Score:3, Interesting)

    by sinij ( 911942 ) on Wednesday December 04, 2013 @09:21PM (#45604225)
    Work by Dr. Charlie Miller showed that in-auto networks have zero security. It wasn't a problem up to this point because such networks were secured by air-gap. Unfortunately automakers decided that facebook integration for the car is worthwhile feature and decided to open Pandora's box. If you are planning to buy a new car, make sure it has no connectivity capability of any kind. This includes On-Star systems, this definitely includes any kind smartphone integration or mobile hotspot technologies.

    Car's CAN Bus is ring network with no authentication whatsoever and rudimentary priority system. If you can broadcast into it, then you can affect operation of the car in very drastic ways. Since it has to be real-time and responsive (e.g. controlling engine timing) there is no time for any kind of authentication. Insanity is allowing things like Entertainment/Navigation/OnStar system access to it, but this is how auto engineers do it. Why? Because they don't know any better, they are not IT Security guys.
    • So you get Pandora in your car, but you open Pandora's box. What's more important?

    • Re:Automotive networks have zero security (Score:5, Insightful)

      by jklovanc ( 1603149 ) on Thursday December 05, 2013 @12:52AM (#45605275)

      Car's CAN Bus is ring network with no authentication whatsoever and rudimentary priority system. If you can broadcast into it, then you can affect operation of the car in very drastic ways.

      Much in the same way as the PCI bus on your computer has "no authentication whatsoever and rudimentary priority system". The bus does not need to be secured. The entry points to the bus need to be secures much in the same way as the Ethernet card provides secure access to the PCI bus.

      Security researchers have taken control of in-auto networks by plugging hardware into the bus. You can do a lot to control a car if you can plug onto the diagnostic port and have a laptop sitting on the passenger seat. I think most people would notice that and be a bit suspicious. There has yet to be a wireless access into an unaltered in-auto system. If that starts to happen then worry.

      Insanity is allowing things like Entertainment/Navigation/OnStar system access to it,

      If the OnStar system is secured and only responds to a specific set of commands I see no "insanity". The whole CAN bus API would not be exposed through the OnStar API. I used to work a a company that facilitated disabling vehicles and locking their doors (It was an application designed for an exotic car rental company. They wanted to be able to disable the vehicle if the vehicle was miss-used). Through our API those were the only commands available. There was no way a hacker could do anything else. The connection to the vehicle was authenticated and encrypted. Every entertainment system has authentication if it uses Bluetooth as authentication is built into the Bluetooth pairing protocol.

      Authentication on the bus is not an issue; authentication at entry points is.

      • Re: (Score:3)

        by AmiMoJo ( 196126 ) *

        The problem is that it is now quite common for the main dashboard computer system that controls the radio/CD player, air con and sat nav to be attached to the CAN bus. If you can hack that remotely and run arbitrary code you could gain access to the CAN bus.

        It's all very well saying that things like OnSat have a "secure" API with a specific set of commands, but what about exploits? Many cars have GSM modems for internet connectivity that contain complete TCP/IP stacks. What if the parser that downloads weat

        • Re: (Score:2)

          by sinij ( 911942 )
          The question is not "what if there was a vulnerability", the question is when people realize these systems are out there and start looking. The answer is - now. We had similar conversations about Internet during 90s, this is the same thing, only now we know how it all turned out.

          Have you followed Toyota runaway acceleration court case? They were forced by courts to undergo low-level code analysis for Toyota's control software. Verdict was that code was unmaintainable. I think they lost the case because

      • Re: (Score:2)

        by sinij ( 911942 )

        >>>Much in the same way as the PCI bus on your computer has "no authentication whatsoever and rudimentary priority system". The bus does not need to be secured.

        Absolutely correct. CAN bus itself does not need to have internal security mechanism - it just has to be properly isolated. By an air gap.

        >>>The entry points to the bus need to be secures much in the same way as the Ethernet card provides secure access to the PCI bus.

        I don't think I have to expl

    • Sounds almost laughably, inevitably fixable.

  • best application: the elderly (Score:3)

    by Gravis Zero ( 934156 ) on Wednesday December 04, 2013 @09:36PM (#45604323)

    at age 65, fatal accidents go waaay up. [dot.gov] i blame the old people sunglasses [cardinal.com] and old people. [grumpyoldsod.com]

  • Imagine how the control of a car will become very accurate when the self-driving part is integrated to the existing computerized parts of a car (stability control and whatnot) and all the components can extremely quickly adapt to the conditions reported by each other.
  • if it's good for them, it's good for you.
  • "The container holding cobalt was found about a kilometer away from the truck and had been opened, he said." ..."At around 1 a.m. Monday, a man armed with a handgun knocked on the passenger window. When the passenger rolled down his window, the gunman demanded the keys to the vehicle, Morales said. Both the driver and his assistant were taken to an empty lot where they were bound and told not to move. They heard one of the assailants use a walkie-talkie type device or phone to tell someone, "It's done," Mor
    • oops -- please disregard, I meant to post in another thread. I'm trying to find a way to remove a post once made, but I can't seem to find it.

  • Given the following:
    http://yro.slashdot.org/story/13/12/04/1817227/ev-owner-arrested-over-5-cents-worth-of-electricity-from-schools-outlet [slashdot.org]
    what would happen if the car decided to recharge itself? Would the car be arrested?

  • All the highway autonomous vehicle projects got as far as freeway driving. BMW, Cadillac, and Mercedes have all demonstrated this level of automatic driving. Then it gets hard.

    This is about as far as you can go before entering the "deadly valley", where the vehicle can drive autonomously but isn't smart enough to recognize when it shoudn't. Google is further along; they can drive around on suburban streets.

    • Re: (Score:2)

      by Qzukk ( 229616 )

      Freeway driving is trivial: you don't hit what's in front of you, you don't hit what's beside you. Basic sensors can pull off both of these feats. You get bonus points if you can stay in a lane, but plenty of shitty human drivers manage to pull it off following those two basic rules.

      Getting off the freeway is where it starts getting difficult. Even google maps sometimes misses the exit and tells me to turn right while I'm doing 60 over an overpass.

      • Re: (Score:2)

        by m00sh ( 2538182 )

        Freeway driving is trivial: you don't hit what's in front of you, you don't hit what's beside you. Basic sensors can pull off both of these feats. You get bonus points if you can stay in a lane, but plenty of shitty human drivers manage to pull it off following those two basic rules.

        Getting off the freeway is where it starts getting difficult. Even google maps sometimes misses the exit and tells me to turn right while I'm doing 60 over an overpass.

        Freeway travel everywhere is pretty much the same.

        On local streets, there are quirks.

        However, this problem can also be solved by doing a Google street view style of predetermined intentions of how the roads were designed instead of computing them on the fly.

        The traffic signals are also not standardized. The yellow in a 45mph road going downhill is shorter than the yellow at a 20mph road. Also, this behavior changes with time of day in some lights. So, when the light turns yellow, the car has to make a

        • You wouldn't need to record light timings. Upon yellow light, determine if safe stop is possible based on distance to light and current speed. If yes, then do so.

          • Re: (Score:2)

            by m00sh ( 2538182 )

            You wouldn't need to record light timings. Upon yellow light, determine if safe stop is possible based on distance to light and current speed. If yes, then do so.

            There will always be a grey area in the decision between if it is safe to stop or not.

            Whatever the condition to determine if it is safe or not, there will always be situations when it will be close the decision boundary.

            Out of the millions stop light encounters, there will be thousands where the decision on either side could be taken depending on the variability of measurements. Two cars taking two different decisions because of minute variations in their sensor measurements could result in a crash.

            If

  • Will it not be possible to "herd" an autonomous car, forcing it in different directions simply by driving up very close to it, triggering it to steer away from the approaching object (that is you in your car). If you and your friend sit in two cars, it will even be quite easy I guess. Imagine how annoying that would be to the passengers of the autonomous car!

    • Well, the passenger of the autonomous car will have both hands free to hold and point that 12-gauge at your tires. Have fun!

    • Re: (Score:2)

      by neminem ( 561346 )

      Because you totally can't force a car being driven by a human in a direction you want just by having cars on either side of it and driving them close to your target. Oh wait, yes you totally can. At least this would be safer - a prankster pulling that sort of thing on a human-driven car would probably have a better-than-even chance of causing a car accident.

  • bad journalism (Score:4, Insightful)

    by Tom ( 822 ) on Thursday December 05, 2013 @04:14AM (#45605987) Homepage Journal

    who no doubt cleared traffic to make the test a little easier.

    Nothing in the article nor in the video backs up this assumption. So why was it in the summary? Having been to Japan, I doubt they would've done this, as the whole point of running the test on a public highway is to show it can cope with other traffic and real-life conditions, and making the test invalid in such a stupid and public way would mean quite a bit of lost face.

Slashdot Top Deals

If you steal from one author it's plagiarism; if you steal from many it's research. -- Wilson Mizner

Close