By on August 25, 2014

1024px-Jurvetson_Google_driverless_car_trimmed

In the entire time Google has been working on delivering an autonomous future upon the driving populace, only one accident was reported, and was caused by human error. That said, the tech giant would prefer you not to know that or of any similar future incidents.

Through a FOIA request, Quartz reports Google lobbied California’s state government for amendments to traffic safety legislation that would, in the words of the company’s director of safety Ron Medford, “limit required reporting to accidents involving vehicles operated in autonomous mode.” The company also wanted language removed that obligated it to report “disengagements” — when the car returns control to the driver — to the government, citing lack of relevance regarding vehicle safety.

Though the state’s DMV wants this data for future driver testing when the day of the autonomous vehicle comes, Medford claimed the reporting would create “a significant burden on manufacturers” and the DMV, especially since the agency did not have “the engineering expertise to interpret the data.”

Other concerns Google wanted the state legislature to address included the fear of trade secrets going public — and to its rivals — as well as the possibility for test drivers to simply leave all of the driving to the car than to disengage even once. The company found support from General Motors, Chrysler, Volkswagen and Mercedes-Benz, all wondering what the fuss was about regarding autonomous vehicle testing and reporting obligations.

The California DMV did excuse manufacturers from reporting every disengagement — only those linked to safety reasons will now be reported to the agency — but did want everything else related to autonomous driving. Google, for its part, was disappointed “that the vast majority” of its comments were ignored.

Get the latest TTAC e-Newsletter!

20 Comments on “Google Wanted Autonomous Vehicle Driver Interventions Kept Quiet...”


  • avatar
    Fred

    When ever I read stories like this I’m reminded of all those scifi stories of computers & robots taking over. It all starts so innocently.

    • 0 avatar
      koshchei

      Hate to be the bearer of bad news, but that particular horse left the barn about 20 years ago.

      In other news, Americans are still scared of the things they invent.

    • 0 avatar
      econobiker

      There are already “autonomous” or robotically guided / controlled vehicles in use TODAY with no drivers on board.

      It is just that these are huge dump trucks in open pit mines in the middle of no where with nothing to hit other than objects already owned by the mining company.

      Reference Caterpillar “Command for Mining” on Youtube for background information

  • avatar
    segfault

    I’m curious as to what kind of MPG the autonomous vehicles get, compared to their conventional brethren. On the one hand, the computer can probably modulate the steering/brakes/throttle more smoothly than a human, and on the other hand, you have the extra weight of the computer equipment and the bad aerodynamic effects of that laser scanner on the roof.

    • 0 avatar
      mcs

      The base autonomous vehicle platform I may be getting soon uses two scanners mounted below the bumper at the ends, so the aerodynamics aren’t as big of an issue. I’m not a huge fan of the scanners, but will take all the data I can get.

      A bigger drag on fuel economy than the weight of the equipment is the drag induced on the alternator by all of the NVidia GPUs sucking up all kinds of power.

  • avatar
    doublechili

    “Disengagements”?!

    Gee, I wonder whether those happen when you’re cruising down the highway with no one nearby, or if it’s more likely to happen in a tight spot? Is this the “autonomous car” equivalent of the panicked driver screaming and throwing both hands up in the air?

    And what will the contractual terms be on that? Do you accept responsibility for “disengagements” in your Agreement? IOW, maybe you won’t actually be able to put your feet up with the smartphone in one hand and a coffee in the other. :)

    Relinquishment of human autonomy does have its funny side after all.

    • 0 avatar
      redrum

      I would imagine there would be a loud warning if auto-pilot (for lack of a better term) were disengaged…I mean, they could just re-use the annoying seat belt reminder chime, it’s loud enough in my car to wake up the deepest sleep.

  • avatar
    thalter

    I can see why Google would not want to report disengagements and accidents with human drivers. Under this scenario, the the car could hand control back to the driver seconds before an imminent collision, and truthfully state that the driver was in control of the car at the time of the crash, and not the computer.

    While such a statement might be technically accurate, it wound not be entirely truthful.

  • avatar
    Carilloskis

    Isn’t the who point of these cars is you ju8st sit in them and enjoy the ride , take a nap read the paper, etc. so if the Car disengages autonomus control randomly then you get none of the perks of an autonomus car. Look at airplanes for example, pilots pay the most attention at take of and landing even with auto land and other advanced systems, they are ready to take controll of the aircraft if they need to, when they are crossing a continent or an ocean there is alot of seperation between them and other aircraft, so if the autopoilot disengeages they have that seperation and 30,000 ft or so to work with. in a car driving alond at 65 in rush hour you dont have that ability to react if your reading a paper or drinking a coffee etc. you most likely havent been paying attentinon to your suroundings and where not expecting the auto pilot to disengage so you end up crashing. They might get these vehicles to avoid hitting other cars , but until they can make decissions and anticipate other peoples actions these systems will cost more lives than they will save. Thke the example of a kid chasing a ball into the street, as a human being we se the ball and slow down the car might decide to avaoid the opbsitcal and hit the child running after it in the process. so then you program that senario in and the car comes to a dead stop when a round object goes infront of it, and the car is driving through an area with tumble weeds and its sensor thinks its spotted a ball.

    There are alot of what if senarios that need to be factored into these vehicles before they go out on public highways, these require inteligence something the machines dont have yet.

    • 0 avatar
      mcs

      My son’s current research involves “rewiring” simple organisms and seeing how it affects their behavior. If that research progresses, we could see autonomous vehicle implementations in what we euphemistically refer to as “post silicon” technology. Artificially grown brains using rat brain neurons.
      http://mcgovern.mit.edu/news/news/precisely-engineering-3-d-brain-tissues/

  • avatar
    jdash1972

    Disengagements are no problem. Every once in a while the CPU freezes or the car thinks it’s in Paris France instead of Paris Texas but it’s a simple fix. You pull over, pop the hood, disconnect the battery for one minute, then reconnect it while holding down the horn button until it boots back up. It only happens a couple of times a day, what’s the big deal? I mean ask yourself, really, when was the last time your GPS took you to the wrong address? It hardly ever happens…., what could possibly go wrong?

  • avatar
    wmba

    The main thing Google doesn’t want you to know is that, since Google is banned in China, their map-based autonomous system will not be deployed in the largest vehicle market. Try getting Streetview for Shanghai.

    The Mercedes system doesn’t rely on mapping everything, or so I’ve read. There’s more than one way to skin a cat, and I see no reason to believe Google’s system is the best, nor do I trust them.

    Hiding autonomous vehicle accidents from the law, based on Google’s interpretation, is no different from Toyota not reporting unintended acceleration events until their engineers spent three years trying to decide if it was important or not, or GM covering up their ignition switches’ failures.

    Is there any sane reason why the fox gets to set the rules in the henhouse?

    That’s why NHTSA seems to have woken up recently and made automakers toe the line on the law. California should set its own rules and make all these jokers play the intended game.

  • avatar
    Broo

    I’ve been working in IT long enough to know I wouldn’t willingly put my life in the “hands” of a software controlled device.

  • avatar
    ttacgreg

    Methinks Google has done too many drugs at Burning Man. Yeah right Google, I want to get into my car and nap to my destination. Not.
    The idea of an autonomous car is ridiculous, even assuming they could achieve what looks to be an impossible goal.
    The legalities and liabilities issues are equally impossible.
    The idea of a cop pulling over a driver who is literally sleeping is an amusing scenario to contemplate.


Back to TopLeave a Reply

You must be logged in to post a comment.

Subscribe without commenting

Recent Comments

New Car Research

Get a Free Dealer Quote

Staff

  • Contributing Writers

  • Jack Baruth, United States
  • Brendan McAleer, Canada
  • Marcelo De Vasconcellos, Brazil
  • Vojta Dobes, Czech Republic
  • Matthias Gasnier, Australia
  • W. Christian 'Mental' Ward, Abu Dhabi
  • Mark Stevenson, Canada
  • Cameron Aubernon, United States
  • J Emerson, United States