Forgot your password?
typodupeerror
Transportation Google

Tesla's Elon Musk Talks With Google About Self-Driving Cars 199

Posted by Soulskill
from the car,-take-me-home-then-find-an-EV-hater-to-annoy dept.
Tesla Motors CEO Elon Musk has been thinking about bringing autonomous driving technology to Tesla's electric cars. Quoting Bloomberg: "Musk, 41, said technologies that can take over for drivers are a logical step in the evolution of cars. He has talked with Google about the self-driving technology it’s been developing, though he prefers to think of applications that are more like an airplane’s autopilot system. 'I like the word autopilot more than I like the word self- driving,' Musk said in an interview. 'Self-driving sounds like it’s going to do something you don’t want it to do. Autopilot is a good thing to have in planes, and we should have it in cars.' ... Google’s approach builds on a push for the driverless-car technology long pursued by the U.S. military’s Defense Advanced Research Projects Agency, which held vehicle competitions for carmakers and research labs. Anthony Levandowski, product manager for Google’s self-driving car project, has said the company expects to release the technology within five years. 'The problem with Google’s current approach is that the sensor system is too expensive,' Musk said. 'It’s better to have an optical system, basically cameras with software that is able to figure out what’s going on just by looking at things.' ... 'I think Tesla will most likely develop its own autopilot system for the car, as I think it should be camera-based, not Lidar-based,' Musk said yesterday in an e-mail. 'However, it is also possible that we do something jointly with Google.'" Musk later warned not to take this as an actual announcement.
This discussion has been archived. No new comments can be posted.

Tesla's Elon Musk Talks With Google About Self-Driving Cars

Comments Filter:
  • IIRC the Grand Challenge winner use a computer vision system, augmented with LIDAR because computer vision is still an evolving field with plenty of risk. I am excited, autonomous navigation in cars seems like the sort of thing that is actually achievable without some major tech breakthrough. Sure it's too expensive currently, but the costs will come down as engineers optimize the system.
    • by Animats (122034)

      IIRC the Grand Challenge winner use a computer vision system, augmented with LIDAR because computer vision is still an evolving field with plenty of risk.

      No, LIDAR works fine, but there's a range problem because of the power limits needed for eye safety. For a nanosecond, a pulsed LIDAR must far outshine the sun.

  • While cameras may be more cost effective than Lidar, they have problems that lidar doesn't. For example, what does the camera sensor do when it's under direct sunlight and can't make sense of what it's seeing? What about rain / fog? I have a feeling google is has the right idea here.

    • by OzPeter (195038)

      While cameras may be more cost effective than Lidar, they have problems that lidar doesn't. For example, what does the camera sensor do when it's under direct sunlight and can't make sense of what it's seeing? What about rain / fog? I have a feeling google is has the right idea here.

      What do *you* do when you're driving and there is direct sunlight in your eyes, or you encounter rain/fog?

      If you could get the software to be half as near as good as peoples eyes and simply take peoples stupidity out of the equation then you'd end up with a system that was miles ahead of manual driving.

      • Yes, well, human eyes have around a million times more dynamic range than a digital camera (about 20 stops vs. 10).

        So the answer is: your eye compensates via a number of mechanisms not available to cameras.

        So "[getting] to be half as near as good as peoples [sic] eyes" will require an improvement in cameras of about 500,000 times.

        • by Arrepiadd (688829)

          Clearly you've never seen what humans do on highways in foggy days...

          People will drive recklessly when they can't see crap. They'll drive too fast, out of a sense of security, out of stupidity, out of an illusion effect coming out of the foggy conditions. And you... well, you can't do much about it, since you only control one gas pedals over the entire highway.

          Computers, on the other hand, can "see" in different wave lengths than we do. Potentially wavelengths that are visible through fog. So, when you don'

    • by drinkypoo (153816)

      I half agree with you. The right approach is probably a combination of millimeter-wave radar, lidar, and computer vision. Not only can you cross-check between systems but at least one of them is sure* to work.

      * for values of sure reasonably near 1.0

  • Lidar might be expensive, but it gives you the shape and depth of the surrounding environment. Camera based imaging will have a harder time determining the distance to the objects in views. I would think the lidar would also have an advantage with fog or rain that might hinder a camera based system much more. In the end I think having multiple systems that corroborate their view of the world and cover for each other when one has difficulty getting a good sense of the environment is the best way to go. But i

  • Major problem here (Score:5, Interesting)

    by vikingpower (768921) <exercitussolus@NOspaM.gmail.com> on Tuesday May 07, 2013 @02:15PM (#43657267) Homepage Journal
    The problem is called "humans". Humans love to bask in the feeling of being in control, especially when it comes to cars. With planes, this was different, especially as these from their beginnings on were called "flying machines", i.e. machines made to fly ( with ). I remember that my grandma, born in 1900, never ever called them differently. Cars, OTOH, have never been called "driving machines". And this is where the crux is hidden: humans want to control their cars. I guess it will remain so for a long time.
    • by mrsquid0 (1335303) on Tuesday May 07, 2013 @02:18PM (#43657299) Homepage

      Some humans want control of their car, but many would rather do other things. The idea of automobiles being an extension of the driver did not really develop until car companies started advertising vehicles that way in the 1950s. Before that automobiles tended to be seen as either luxury items or utility vehicles.

    • by i kan reed (749298) on Tuesday May 07, 2013 @02:24PM (#43657367) Homepage Journal

      Cars, OTOH, have never been called "driving machines".

      Remind me again what BMW's tagline is?

    • Cars, OTOH, have never been called "driving machines".

      I take it you've never seen a BMW ad; their main tagline is (and has been, for as long as I can remember*) "BMW: The Ultimate Driving Machine"

      * I wasn't around way back when they were making engines for the Luftewaffe.

    • by steelfood (895457)

      The lack of control in an airplane is one of the reasons why the newest "anti-"terrorism security measures are so easy to implement in the airport.

      People want control. Period.

    • by dkleinsc (563838) on Tuesday May 07, 2013 @02:46PM (#43657619) Homepage

      Humans love to bask in the feeling of being in control, especially when it comes to cars.

      Specifically, there's an important cognitive bias at work here, in that people feel safer about things they control than about things they don't control. That's why people who feel perfectly safe driving feel unsafe riding a commercial aircraft, even though planes are much much safer than driving. That's also why geeks feel comfortable with computers, while non-geeks are frequently scared of them - geeks know how to control those machines, non-geeks don't.

      • by drinkypoo (153816)

        That's why people who feel perfectly safe driving feel unsafe riding a commercial aircraft, even though planes are much much safer than driving.

        Not that I drive as safely as I potentially could, but in the case of plane it's sit back and hope the pilot is competent, whereas when driving there's lots of things you can do to mitigate the potential danger like driving defensively (I do that, anyway) and driving slowly and driving a vehicle which does well in crash tests and taking breaks etc etc. (too many ands, too lazy to edit.) You can choose your routes as well, within parameters. So even if it never gets as safe as a plane on average, you still h

    • by aztektum (170569)

      I think it's more we're conditioned to feel that way about our cars, thanks to decades of advertising designed to do so.

      The reality is that we have very little control over our driving. Collectively we're spending billions of hours each year stuck in traffic. We burn billions of gallons of gas going nowhere.

      Replace that with a largely automated system that can route around traffic issues, reduces the number of cars needed on the road, and you actually return control to folks.

      Plus cars are old technology. Th

    • Humans love to bask in the feeling of being in control, especially when it comes to cars. With planes, this was different

      Haven't met many pilots, have you?

    • Humans love to bask in the feeling of being in control, especially when it comes to cars.

      You forgot about buses, taxis, and also ordinary passengers in your own car. People will happily cede control to a driver they trust.

      I enjoy driving as much as the next guy, but not all the time, not when I'm too tired/drunk, and rarely in traffic. A significant majority of driving is done for utility, not enjoyment. Personally I would welcome a car with the option of an autopilot.

  • Er, TFA? (Score:5, Funny)

    by raddan (519638) * on Tuesday May 07, 2013 @02:18PM (#43657297)
    I know we're probably not going to read the articles, but... can't we have a link just for old time's sake?
  • by RandCraw (1047302) on Tuesday May 07, 2013 @02:28PM (#43657415)
    Tesla CEO Talking with Google About 'Autopilot' Systems

    http://www.bloomberg.com/news/2013-05-07/tesla-ceo-talking-with-google-about-autopilot-systems.html [bloomberg.com]

  • by Animats (122034) on Tuesday May 07, 2013 @02:40PM (#43657555) Homepage

    As someone who's actually done this stuff, LIDAR gives solid data, but is range-limited. Cameras have more ambiguous results. Cameras are most useful when things are going well, as on a highway under good conditions. That was Stanford's approach in the Grand Challenge. All their vision system really did was answer the question "is the near section of road (within LIDAR range) like the far section of road"? If the LIDAR said the near section was OK to drive on and the vision system said the far section was like the near section, then the vehicle could speed up and out-drive the LIDAR range. That sped up travel on good sections of road.

    Google is using Velodyne LIDAR units, which are effective but an expensive mechanical kludge. A better approach is from Advanced Scientific Concepts [advancedsc...ncepts.com], which has an eye-safe flash LIDAR. No moving parts.

    ASC's units cost about $100K each, but that's because they're hand-made for DoD. The technology isn't inherently expensive if made in volume. It uses custom imaging ICs, and because they're made by tens, not millions, they cost far too much. If the cost can be brought down, the vehicle can have multiple LIDAR units around the car to get full coverage, rather than one big spinning thing up on the roof.

    Millimeter radar is also useful. It's good to have a Dopper anticollision radar as a backup system. It provides an unambiguous "rapidly approaching big solid object" signal. We had one of those on our DARPA Grand Challenge vehicle as a backup to the fancier LIDAR system.

  • by Bob9113 (14996) on Tuesday May 07, 2013 @02:53PM (#43657687) Homepage

    Autopilot is a good thing to have in planes, and we should have it in cars.

    I like the notion, and it's a great frame of reference for consideration. One major distinction between planes and cars: When a plane is on autopilot in a relatively sparse chunk of sky, the time between sensor warning and twisted burning wreckage is tens of seconds to minutes. Most of the time in an ordinary flight plan the plane can wander hundreds of feet without a problem. On a typical chunk of sparsely populated two lane highway, however, If your car's autopilot travels twenty feet out of its lane -- things get exciting very quickly.

    Moreover, most airplanes are like long-haul trucks -- they spend most of their miles in transit between heavy traffic areas. A major chunk of American automotive miles are spent with other vehicles within a few dozen feet.

    • Also, all planes have some at least 2 highly trained professionals sitting in the cabin overseeing everything. People who have to log a minimum number of flight hours in order to keep their skills sharp. In the event of a pre "burning wreckage" moment, they can take control if necessary.

      I can't see your average car driver (who probably is not paying attention on the road) do the same, especially if they have gotten a bit rusty from constantly allowing the car to drive for them.

      Autopilots for cars and planes

      • by tlhIngan (30335)

        Also, all planes have some at least 2 highly trained professionals sitting in the cabin overseeing everything. People who have to log a minimum number of flight hours in order to keep their skills sharp. In the event of a pre "burning wreckage" moment, they can take control if necessary.

        False. Many planes require only one crewmember - the pilot. This is true whether the pilot is flying for himself and friends (e.g., private pilot) or doing it as part of a charter (air taxi).

        And even the dinky little Cessnas

  • Insurance (Score:5, Interesting)

    by EmperorOfCanada (1332175) on Tuesday May 07, 2013 @04:05PM (#43658393)
    Once cars are robotic enough that we are only giving guidance to the car (pretending to drive but the car will ignore stupid inputs) insurance will become the battleground. I can see a car company like Tesla offering free liability insurance with their cars. They will know that basically their car can not cause an accident and with the camera/computer system will have the proof to avoid a he-said-she-said situation.

    At the same time I can see the insurance companies realizing that a huge huge HUGE market will simply go away when car accidents become unlikely enough for car companies to be able to cover it. Think about it. Every car that you see is paying in around $1,000+ for insurance. The only insurance people will want after robotic cars will be theft (hard to do with a hi-tech upgradable car), vandalism, trees falling on them kinds of insurance. Plus nearly every jurisdiction says you must have something like 2 million in liability; that need will vanish or at least be covered by the manufacturers.

    So my robotic car prediction is that car companies will be trying to terrorize us into hating robotic cars. They will show videos of families being driven off cliffs, or saying it is our god given right to have control of our cars. And of course they will spend ungodly amounts of money lobbying everyone from the president down to your school board to stop this.

    But the simple reality is that 35,000 people are killed every year in the US and robotic cars might take this down to a few hundred. (mechanical failure, trees falling on them, sinkholes, etc)
    • So my robotic car prediction is that car companies will be trying to terrorize us into hating robotic cars.

      Car companies or insurance companies?

      • Insurance companies. Car companies will love it if all non robotic cars are booted off the roads as weapons of mass destruction. Gear heads will cry and wail but 35,000 deaths is a 9/11 every month.
    • criminal liability

      Who pays a ticket? Who get's points? Can you have a bot take traffic school?

      Now on to bigger stuff What it does some thing that you can get jail time / big fines / ect...

      What it drives into some one at speed and does not even to try to stop? misses a road closed and hits people in the street?

      Hit's a work worker? Will Mushishkabishalishdish Jaboodiboodi do the 14 years and pay the 10K fine? Will all people get off how fast will that EULA be cut down in a court? will Google take the 5th?

  • Often I sight-see, get intentionally lost, or even prefer to visit locations with no name (to hike, to consider buying, etc).
    Are driving cars another round of "consumer" vs "creator"?

Did you know that for the price of a 280-Z you can buy two Z-80's? -- P.J. Plauger

Working...