Slashdot Mirror


Road Train Completes First Trials In Sweden

Hugh Pickens writes "BBC reports technology that links vehicles into 'road trains' that can travel as a semi-autonomous convoy has undergone its first real world tests with trials held on Volvo's test track in Sweden. Researchers believe platoons of cars could be traveling on Europe's roads within a decade cutting fuel use, boosting safety and may even reducing congestion. SARTRE researchers say that around 80% of accidents on the road are due to human error so using professional lead drivers to take the strain on long journeys could, they say, see road accidents fall. They also predict fuel efficiency could improve by as much as 20% if 'vehicle platooning' takes off, with obvious benefits for the environment. 'An automated system is likely to make it safer as it takes away driver error but it would have to be 100% reliable,' says John Franklin 'This kind of system would also require a complete change in motoring culture for drivers to hand over control.'"

30 of 345 comments (clear)

  1. 80% due to human error? by somersault · · Score: 3, Interesting

    That figure seems a bit low. Unless an animal runs across the road or similar, other problems are all IMO human error.

    If something falls off a truck, that's human error for not securing the load properly. If high winds knock over your truck, that's human error for driving in dangerous conditions. If you skid on an ice patch, that's your error for driving too fast for the conditions, etc.

    --
    which is totally what she said
    1. Re:80% due to human error? by orphiuchus · · Score: 4, Informative

      Tire blowouts, serpentine belt breaks, break cylinders exploding. Sometimes its a maintenance issue, but a lot of the time things just fail. I had the rear cylinder explode on a vehicle I was driving at highway speeds like 10 years ago, I just barely managed to stop myself with the hand-break on the shoulder without running into the stopped traffic ahead of me, but if I had hit then 100% of my accidents ever would have been caused by mechanical failure with no forewarning.

    2. Re:80% due to human error? by somersault · · Score: 2

      For high speed accidents then unexpected failures would probably count for more (though I still think that 1 in 5 is giving people too much credit). Even at lower speeds though, there's still serious danger of maiming or killing pedestrians, so even those women on their cell phones are a danger.

      I was taught that basically all accidents are human error. This page claims it's at least 95%. Too many people try to blame external factors when in fact the accident was avoidable. I really don't like to hear that someone crashed "because it was raining/icy/snowy". They crashed because they were driving too fast for the conditions.

      --
      which is totally what she said
    3. Re:80% due to human error? by Kjella · · Score: 3, Insightful

      I was taught that basically all accidents are human error. This page claims it's at least 95%. Too many people try to blame external factors when in fact the accident was avoidable. I really don't like to hear that someone crashed "because it was raining/icy/snowy". They crashed because they were driving too fast for the conditions.

      Every accident would be avoidable if you drive at 5 km/hr, no matter what the conditions are. The question is what is reasonable and unreasonable to expect, if you hide behind a tree near a high speed road and jump out in front of a trailer you will with 99% probability get splattered even if it's a perfect day and the driver goes no faster than the posted limit. But in retrospect you can always claim it's human error and too fast, even though that's how fast we actually expect people to go. In fact under good conditions they will fine you for being way below the limit.

      Even if you're driving at speeds that seem reasonable given that it's snowy and icy you can get caught by surprise. I've been off the road once because I got tricked by a bus pocket. It was heavy snowfall, I was already going something like 50 km/h instead of the limit of 80 km/h and for the briefest of moments I followed the curve into the pocket. The road was quite well trafficked and worn clear, but in the pocket there was nothing but polished ice with light snow on top so nothing could get a grip. I couldn't steer, couldn't brake fast enough and went off the road at the end of the pocket. I checked now in a calculator and I couldn't have stopped on 15 meters of ice with with 30 km/h (20 mph) and one second reaction time.

      I suppose you could call it human error. But either you assume I would have avoided the situation - which is unlikely - or it would really taken a massive speed reduction to avoid it. Like way, way below what people normally drive, even under those conditions. Either that makes 95% of us reckless or it's jusr acknowledging that driving that car at those speeds under those circumstances is an acceptable but non-zero risk for all the benefits and liberty it gives drivers and their passengers. Not that we shouldn't make roads and cars safer, but until something will literally block me from driving over a pedestrian or off the road we will have accidents.

      --
      Live today, because you never know what tomorrow brings
    4. Re:80% due to human error? by GooberToo · · Score: 2

      so even those women on their cell phones are a danger.

      A fact which always surprises people is that some stereotypes are actually true. Women are far, far worse drivers then men. BUT, while women have far more frequent wrecks, they are typically less severe than those created by male drivers. Whereas male drivers have far less frequent accidents but when they do wreck its typically far more serious. This is why young males are more expensive to insure.

      So if you say women are bad drivers, its not only a stereotype but statistically accurate. Just the same, a male driver is far more likely to kill or seriously injure you.

    5. Re:80% due to human error? by pixelpusher220 · · Score: 3, Insightful

      Every accident would be avoidable if you drive at 5 km/hr, no matter what the conditions are.

      Never dealt with ice have you? ;-)

      --
      People in cars cause accidents....accidents in cars cause people :-D
    6. Re:80% due to human error? by GooberToo · · Score: 3, Insightful

      Seems to me, you've universally redefined what a bad driver is so as to have an irrational rant, which adds absolutely nothing to the thread at hand. I don't believe you'll find any reasonable person (which seemingly excludes you) who will argue that a bad driver is not someone who has lots of wrecks. Inversely, you're not going to find a reasonable person (again, excluding you) who will argue the definition of a good driver is one who has frequent wrecks.

      Made even worse is the fact that your completely irrational and unique definition means all of the world's top drivers are, according to you, "bad drivers."

      I'm sorry, but your post is ignorant to say the least.

    7. Re:80% due to human error? by JonySuede · · Score: 2

      Which driver will have its driving license taken away?

      In a sane world, both.

      --
      Jehovah be praised, Oracle was not selected
  2. This is so 1970s by Kupfernigk · · Score: 2
    An engineering lecturer at Cambridge was proposing something like this in the early 1970s, but with vehicles having a mechanical connection - inherently safer because sudden braking would merely load up a damper, not cause an impact.

    The problem where the UK is concerned is that motorways are actually our safest roads - it's people like the idiot woman this morning in the Range Rover who think that size overrides the Highway Code that present the problem, and this doesn't address it.

    --
    From scarped cliff or quarried stone she cries "A thousand types are gone, I care for nothing, no not one."
    1. Re:This is so 1970s by Chrisq · · Score: 2

      I am sure it will be even better if you drive 40mph and even better with 30, 20, ... all the way to zero. So why not drive 30mph? I am sure 2 wheels will be more fuel efficient as well. So take a moped and drive 10mph if fuel efficiency is what is important to you.

      No, nearly all cars drop in fuel efficiency below 20mph. I can't find more recent information but it looks like many cars are most fuel efficient at somewhere between 30 and 55 mph, have a quite flat consumption in that range, and drop off below this fairly quickly and above this rather more slowly.

  3. Never going to work in a litigious society by orphiuchus · · Score: 5, Insightful

    One of my engineering professors worked on something like this in the 90s, when I asked him why we never saw anything like this come into use he said that they knew that the first time anyone was killed in an accident involving one of the automated vehicles the entire project would be dead. Regardless of if it was from something like a blowout causing a computer driven car to swerve into the other lane, or some drunk ramming headlong into a "platoon" of cars.

    Even if it is much safer, the lawyers will be salivating while they wait for the first death.

    1. Re:Never going to work in a litigious society by teh+kurisu · · Score: 4, Insightful

      The protocol for leaving the platoon and returning manual control to the driver is going to be the most difficult thing to solve I think, particularly where it occurs in a emergency situation. A blowout on the motorway is dangerous enough, but a blowout on the motorway where control of a car is suddenly returned to a driver in the middle of drinking a coffee and reading a newspaper could be disastrous.

    2. Re:Never going to work in a litigious society by timholman · · Score: 4, Insightful

      Even if it is much safer, the lawyers will be salivating while they wait for the first death.

      I think the fear of lawsuits preventing autonomous vehicles is way overblown.

      Historically, the auto industry has had several design flaws that have led to huge lawsuits, e.g. exploding gas tanks on the Ford Pinto. Ford's gas tank design led to numerous deaths and injuries, and corporate memos later showed that the company was even aware of the problem, yet Ford was not sued out of business. Even today, with all the fuss and lawsuits concerning Toyota's computer systems, Toyota is doing just fine. Lawsuits are part of the cost of doing business in the auto industry.

      The technology being used in autonomous vehicle research was, by modern standards, painfully primitive 20 or 30 years ago. I could see how people would fear legal liability, because those older systems weren't smart enough to deal with every contingency in a roadway environment. Today's research vehicles are much better, and in ten years they'll be even better still.

      The question to ask is this: can autonomous vehicles do better than 35,000+ fatalities, 2 million+ injuries, and $200B+ in liability / medical costs per year? That's what the U.S. alone is paying right now with humans behind the wheel. 20 years ago, engineers knew their vehicles weren't robust enough for the roadway. As Google's own experiments have recently shown, things are much different now.

      There's no doubt that autonomous vehicles will fail from time to time, and occasionally someone will be injured. But fatalities from a well-engineered system will be rare, and the roadways will be orders of magnitude safer. The fear of autonomous vehicles is basically a classic example of flawed risk perception by human beings - they are uncomfortable with a few hundred possible auto accidents with a computer in control, yet think nothing of millions of accidents with the current system because they all think "I'm in control of the situation".

    3. Re:Never going to work in a litigious society by Max+Romantschuk · · Score: 3, Insightful

      Even if it is much safer, the lawyers will be salivating while they wait for the first death.

      There is a good reason why this project is sponsored by the EU and not the US.

      While not perfect, the legal systems in most of Europe aren't not quite as broken as in the US.

      --
      .: Max Romantschuk :: http://max.romantschuk.fi/
    4. Re:Never going to work in a litigious society by Anonymous Coward · · Score: 2

      There is a reason US innovation has begun a slow march backward. You'd have to be out of your gourd to bring anything new to market in the States, these days.

    5. Re:Never going to work in a litigious society by Chrisq · · Score: 4, Insightful

      The question to ask is this: can autonomous vehicles do better than 35,000+ fatalities, 2 million+ injuries, and $200B+ in liability / medical costs per year? That's what the U.S. alone is paying right now with humans behind the wheel

      That is the sensible question, but in reality it would have to be much safer to be accepted. We see this when there are train crashes. A train is already hundreds of times safer than a car but there are public inquiries, people brought to book and calls for improvement when they happen.

    6. Re:Never going to work in a litigious society by hypertex · · Score: 4, Interesting

      Even if it is much safer, the lawyers will be salivating while they wait for the first death.

      There is a good reason why this project is sponsored by the EU and not the US.

      While not perfect, the legal systems in most of Europe aren't not quite as broken as in the US.

      There were tests of this in California in the 1991 timeframe but I don't know if it was the State or the Feds. A train of 5 white cars would assemble at speed on interstate 15 between Palm Springs and San Diego. With only inches in between, the train would travel back and forth on the freeway . Perhaps another can find a record of this as my Google-fu is not adequate.

    7. Re:Never going to work in a litigious society by adamchou · · Score: 2

      Lets take that blowout idea one step further and imagine the driver in the lead truck somehow losing control, going over the side of a cliff, and a pack full of lemmings following suit. I'm sure that'll be awesome.

  4. Less driver attention == lower safety? by pmontra · · Score: 2, Insightful

    You can trust the system, but the system doesn't know what's happening to your car. It knows what's happening to the leading truck. Suppose that a car in the convoy has a failure, a blown tire, anything that makes it slow down or change trajectory (maybe some bump or hole in the road). How do following cars avoid it if their drivers are sleeping, reading a book, having lunch? I know that people start car accidents while they are driving (texting, playing with music controls, having lunch) but I wonder if road trains are really safer than an equivalent number of cars each with its own driver. I think that this is the only safe road train.

    1. Re:Less driver attention == lower safety? by Anonymusing · · Score: 2

      You can trust the system, but the system doesn't know what's happening to your car. It knows what's happening to the leading truck. Suppose that a car in the convoy has a failure, a blown tire, anything that makes it slow down or change trajectory (maybe some bump or hole in the road). How do following cars avoid it if their drivers are sleeping, reading a book, having lunch?

      Exactly. Or maybe it's not a failure, per se, but something as simple as running out of gas? Is the system going to communicate all this information to the lead driver? Is that driver going to be responsible for alerting individual drivers that they need to jump out of the train to fuel up? Will the train just automatically pull into rest areas/gas stations and have *everyone* fuel up?

      These are not insurmountable questions, but they do suggest a slower adoption rate (or smaller market) for the technology.

      --
      Liberal? Conservative? Compare perspectives at Left-Right
    2. Re:Less driver attention == lower safety? by Tom · · Score: 2

      Suppose that a car in the convoy has a failure, a blown tire, anything that makes it slow down or change trajectory (maybe some bump or hole in the road). How do following cars avoid it if their drivers are sleeping, reading a book, having lunch?

      RTFA. It's even embedded. They do say that they have a system in place monitoring the road, the distance to the car in front, etc. They are not just blindly sending instructions from the lead car and executing them.

      but I wonder if road trains are really safer than an equivalent number of cars each with its own driver.

      Statistically speaking, if the average chance of a driver having an accident is 1% (it isn't, but it's easier to calculate with simple numbers), then 10 individual cars will have a total probability of 9.562% of at least one car having an accident. The road train will have a probability of 1%. The risks associated with driving in a road train (failure of the system, etc.) would have to raise the accident probability by a full order of magnitude to make it more risky.

      --
      Assorted stuff I do sometimes: Lemuria.org
    3. Re:Less driver attention == lower safety? by FatLittleMonkey · · Score: 2

      I know the system's still in early trials, but I can't see anything about safety testing in their own site (or any article about them). They've tested in snow, so I assume they've seen a bit of ice, but there's nothing about them deliberately trying to find the system's fail points.

      And that seems a little too common in these automated systems. Remember Volvo's infamous lolfail video? An anti-collision system that happily drove into a truck. Or Top Gear's experience with a production auto-park system, which happily backed into a fence. These systems work brilliantly until you throw something at them that wasn't in their test.

      How will they cope with faulty data? What happens when I buy one that's 20 years old? I have enough trouble with faulty sensors on my decidedly non-automated car.

      I can imagine people reporting their cars suddenly swerving off the road for no apparent reason. Manufacturer stonewalls, claims driver error, publishes data from in-car data recorder showing that a manual "release" was triggered. Except it is recorded from the same sensor that triggers the release, so can't differentiate between a driver action and a faulty sensor. Then someone dies. And the manufacturer issues a world-wide "fix", a slightly higher surround around the Release button.

      --
      Science is all about firing a drunk pig out of a cannon just to see what happens.
  5. Re:I dream of a day by Eudial · · Score: 2

    Perhaps one day we could have automated platoons of Slashdot submissions about the same damn thing, too?

    The last submission was about SARTRE before the tests started. This is the results.

    --
    GAAH! MY PRINTER IS ON FIRE!!! PUT IT OUT! PUT IT OUT!
  6. Re:Causes vs circumstances by orphiuchus · · Score: 2

    Usually, too much importance is given to the immediate cause of an accident. Most accidents don't happen due to a single cause, there's a number of circumstances that must exist together for an accident to happen.

    In modern highways, the usual circumstance for most accidents is crowded lanes. The usual cause for crowded lanes is a few dumbasses of the i-hate-tailgaters-and-i-have-the-right-to-drive-at-any-speed-below-the-limit species.

    Make it a severe offense, same penalties as drunken driving, to drive on the left lane with someone behind you and those "80% accidents" will go away.

    Make it a severe offense to tailgate and you get the same solution, plus its the assholes instead of the timid that you would be punishing.

    I get it when someone is going 55 in the left lane, that's insanity, but the vast majority of people whom I know that think like this are usually going 15 over the speed limit and riding the ass of the guy going 5 over. The tailgaters and aggressive drivers are the ones who cause fatalities.

  7. Re:I dream of a day by CheerfulMacFanboy · · Score: 2

    Perhaps one day we could have automated platoons of Slashdot submissions about the same damn thing, too?

    The last submission was about SARTRE before the tests started. This is the results.

    Ohh? http://tech.slashdot.org/story/11/01/18/0411235/How-Europe-Will-Lower-Emissions-mdash-Self-Driving-Cars (2 days old)

    The team behind SARTRE has now conducted its first real world test, using a sole Volvo S60 sedan that followed a lead truck around the automaker's test facility near Gothenburg, Sweden. In the video, the driver is free to take his eyes off the road and his hands off the wheel. In fact, he uses neither his hands nor feet during the test.

    And yes, it's the same fucking video. DUPE.

    --
    Fandroids hate facts.
  8. DC Metro, anyone? by Loosifur · · Score: 2

    Good luck selling this to anyone from the Washington, DC area. The Metro is, rightly or wrongly, notorious in the DC area for being dangerous. The WMATA is notorious for everything from ignoring safety recommendations, running old cars, and skipping maintenance, to promoting a culture of hostility within its workforce. Metro employees are underpaid, overworked, and, to put it delicately, benefit from a somewhat lenient hiring process. Now, who would you propose will be driving the lead car around the Capital Beltway? Unless you pick this one segment of public transportation to be contracted out to a private company, it's gonna be the WMATA in the DC area. If I wanted some surly bastard with no professional training who hates his job and hasn't slept in a day to drive, I'd do it myself, thanks.

    --
    This unbiased moderation brought to you by the Porcine Aviation Group!
  9. Re:Causes vs circumstances by Zironic · · Score: 5, Insightful

    You fail at both driving and physics. To get a safe distance to the car infront of you you only need to slow down for 5 seconds at most then you can match speed with car infront of you, there is no recursive slow down for the entire road.

    "It takes two to tailgate, a tailgater and a tailgatee. As I said, accidents usually aren't caused by one single cause, in most cases if just one of a set of circumstances didn't happen there would be no accident."

    That's like saying it takes two to punch someone, the puncher and the punchee. It's retarded. There exists none, zero, zip, nada excuse to tailgate. There exists no situation where you're better off tailgating the person infront of you, you don't even get to your destination faster.

    And for the record it is already illegal in most of the world to drive slow in the left lane (Usually under some law conserning disrupting traffic ) however that only applies if he's driving slower then the limit, if you think that's to slow then that's your problem. However there's never really any reason to go faster, suppose you drive at 110 on a 100 mph road, now your 30 minute trip takes 27 minutes, who cares? Those 3 minutes are a rounding error of your day.

  10. Re:Serial pile-up? by PseudonymousBraveguy · · Score: 2

    Of course, and if the lead driver gets killed in an accident and some of the passangers in other cars survive the crash, they will be automatically shot to follow the lead as closely as possible. Also, everbody who works in automotive research is stupid and not able to see this obvious problem.

  11. Auto Rail by __aazsst3756 · · Score: 2

    What if we built a single rail that mimics the interstate system, but made to hurdle small cars that meet a standard specification for size, weight, and aerodynamic profile. Perhaps the track is a single electromagnetic pipe cars hung from? You could then purchase a car with a built in connector to operate on this track, and a dash panel to choose destination and cost. You can drive local, and hop on the rail to travel longer distances. The rail control computer could slow and speed individual cars to group them for aerodynamic efficiency.

  12. Re:Causes vs circumstances by tragedy · · Score: 2

    I think, given the disagreement on your claim, it would be best if you gave an actual cite that people could actually look up. Not just "it's in this other book", but actually something like "it's in section blah, subsection blahblah of California general law blahblahblah". That would clearly silence your critics if it confirms what you assert, and it would stop all this speculative back and forth.