By on September 5, 2015

Toyota announces $50 million in funding partnership with MIT and Stanford University for artificial intelligence research collaboration.

Toyota announced Friday it would invest $50 million in research facilities at Stanford and the Massachusetts Institute of Technology to study and develop artificial intelligence for future safety and autonomous driving.

The facilities will teach computers to recognize and monitor objects — a swerving car vs. a parking one was provided as one example — on the road that drivers are too busy for because “Candy Crush.”

The joint programs at MIT and Stanford will first develop enhanced safety systems designed to “share control” with drivers and computers. Eventually, researchers believe, people will just forget that they care and give up driving to the robots.

“AI-assisted driving is a perfect platform for advancing fundamental human-centric artificial intelligence research while also producing practical applications,” Fei-Fei Li, an associate professor of computer science at Stanford, director of SAIL and the director of the new AI center, said according to the manufacturer. “Autonomous driving provides a scenario where AI can deliver smart tools for assistance in decision making and planning to human drivers.”

The Stanford lab was created with the Stanford Artificial Intelligence Lab, which helped create chess-playing computers in the 1960s, pioneered AI in the 1970s and recently competed in the 2007 DARPA urban challenge for autonomous cars.

(H/T to David for sending this over.)

Get the latest TTAC e-Newsletter!

Recommended

23 Comments on “Toyota Will Spend $50M Researching The Perfect Robot Car...”


  • avatar

    I hope they give it tentacles and self awareness.

    It’ll get Fei Fei up there FIRST.

    Probably redesign a better body for itself to make itself a faster, more efficient hunter of human females.

  • avatar
    RideHeight

    Does such a relative pittance from Toyota mean they think AVs are the cars of the future and always will be?

    • 0 avatar
      Pch101

      It isn’t much money compared to R&D spending, but it’s a fair amount for a university grant.

      Artificial intelligence is a bit outside the realm of TMC’s usual work. Not surprising that it would outsource this to universities.

      • 0 avatar
        RideHeight

        My first thought was why an American school? Shades of Japan Inc. circa 1985, sponsoring foreign university chairs & scholarships left & right?

        Sadly, you have to go down to #34 on US News’ list of global university rankings to find a Japanese school, so I guess this isn’t just a PR move.

        • 0 avatar
          WheelMcCoy

          “My first thought was why an American school? Shades of Japan Inc. circa 1985,”

          Silicon Valley is still the envy of the world. U.S. degrees — at least in the tech field — rate highly.

          I just wish other countries, and even the East Coast, would stop trying to replicate Silicon Valley. Yes, study it. Use it for inspiration. But to copy it? Meh. They would be better off playing to their own local strengths.

        • 0 avatar
          Frankie the Hollywood Scum

          Interesting, I used to work at a Japanese car company in R&D if we brought up an idea that originated from an non Japanese source it was reflexively dismissed as being “low technique.”

          Of course when I was in Japan they would often have me help translate SAE papers often written from sources outside Japan.

          This strikes me as a pragmatic move by TMC. Better to be ready for a major market move if it happens than have nothing ready to bring to the table. And heck we are only talking about 0.5% of their yearly R&D budget so this is not a all in move.

  • avatar
    APaGttH

    $50 million is couch cushion money for Toyota.

    • 0 avatar
      WheelMcCoy

      “$50 million is couch cushion money for Toyota.”

      Sounds about right. $50 million to send your car to Stanford and MIT to earn a degree in AI. In 10 years, that will probably be how much it will cost to send our kids. :)

  • avatar
    Joss

    The first-ever convincing robot car should be developed in India because they’ve got the traffic situation and the software engineers.

  • avatar
    Volt 230

    They could have used that money to develop a better subcompact car.

  • avatar
    SCE to AUX

    As long as human drivers remain legally responsible for crashes, they won’t buy such vehicles.

  • avatar
    John

    If you use cruise control, you are already “sharing control” with computers – ditto for anti-lock brakes, traction control, launch control, parking assist, etc., etc.. Even the automatic “seek scan” function of your auto’s radio features “shared” human/robot “control”.

    • 0 avatar
      brn

      I’m still telling the car how fast to go, when to accelerate, when to brake, when to go right and when to go left.

      If the car does differently, than there’s a potential lawsuit to the manufacturer.

  • avatar
    shaker

    I’m becoming a Luddite; not because I wish to, but the push to total reliance on technology – driven purely by the profit motive.

    Of course, my tune may change when I can play chess with my car while in traffic on the way to the regeneration facility.

    • 0 avatar
      RideHeight

      Yepper, I jumped off the treadmill when everything went cloud.

      But I’ll jump back on when we have The Internet of IV Pumps!

    • 0 avatar
      WheelMcCoy

      “I’m becoming a Luddite; not because I wish to, but the push to total reliance on technology – driven purely by the profit motive.”

      Agree. Even Google agrees as they have scaled back their ambitions realizing it’s not reasonable to have a human who is out of practice take control of a car in trouble.

      But some other companies will see the profit and not the downsides, and sell the idea to investors who also see nothing but the up side. This will force some companies — competing or after-market support — to jump into the fray prematurely. A disaster will happen. People will die. And then recalls and regulations will follow.

      Such is the way of a free market. I’d prefer a smart market that anticipates disasters.

      • 0 avatar
        akatsuki

        No. You are wrong. The only accidents have been when a human has taken the wheel.

        Take off the tinfoil hat.

        Automakers aren’t dumb, they know that autonomous cars will actually kill them in the long run. When a car is autonomous, a service like Uber will become much, much more credible. Less cars on the road = less profits.

        • 0 avatar
          WheelMcCoy

          >> The only accidents have been when a human has taken the wheel.

          I am well aware of that. And the percentage is even lower than you think because the fender benders involving Google cars were compared to human driven fender benders in police reports. That does not include the countless human driven fender benders that go *unreported.*

          But autonomous cars can’t handle every situation. After a harsh winter, the FDR Drive in NYC was pot-hole ridden. And the lane paint was scraped away by snow plows. And there are underpasses that block or confuse gps and cellular signals.

          The transition period where humans and autonomous cars share the road is just as important as your final ideal state where every car is Uber Autonomous. It deserves an equal amount of fore-thought, if not more.

          • 0 avatar
            shaker

            “Forethought” implies “planning”, which is an anathema to the Free Market.

            Let’s just pitch it at the wall, and see what sticks.

            Oh, and it will be up to the government (taxes) to establish and enforce the (inevitable) regulations regarding this.

            And wait until the insurance companies (who will jump for joy, initially) find out the cost of all the lawsuits due to unintended consequences of these cars – when a hacking will possibly kill someone.

            Edit: I’m ENTIRELY in support of technology that will prevent accidents due to health problems (we’re not getting any younger!) or other driver incapacity, but these have to have a very thick firewall from “The Internet Of Things”. Computers are more powerful than ever, and don’t need a constant connection to “The Cloud” to be functional – that’s just the desire of corporations to increase our reliance on it, and they are pushing the “added value” of it while ignoring the security issues.

  • avatar
    mcs

    Then there’s this little item. Most AV researchers think they’re about to conquer Everest and they aren’t even to the base camp yet.

    http://spectrum.ieee.org/cars-that-think/transportation/self-driving/researcher-hacks-selfdriving-car-sensors

Read all comments

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • Scoutdude: You don’t want to use that wrap stuff on your calipers, it says it is for wheels and accessories and...
  • Scoutdude: Back when we used to turn rotors instead of replace them it wasn’t uncommon to find “Jesus...
  • Peter Gazis: FreedMike The Pilot has more space. Becky wins!
  • Scoutdude: There are some premium calipers out there that are powder coated in the OE color if that was they way the...
  • Jeff Semenak: I bought a 2001 Olds Bravada used, in 2004. The most left Button under the Stereo was unmarked and, the...

New Car Research

Get a Free Dealer Quote

Who We Are

  • Matthew Guy
  • Timothy Cain
  • Adam Tonge
  • Bozi Tatarevic
  • Chris Tonn
  • Corey Lewis
  • Mark Baruth
  • Ronnie Schreiber