Tesla Model S With Manufacturer’s Plate Caught Testing Velodyne Lidar System – Video

10 months ago by Eric Loveday 23

Testing Lidar

Testing Lidar

While Tesla continues to insist that it won’t switch over to LIDAR for its sensor system for Autopilot, the sightings of various manufacturer plate Tesla Model S electric cars seems to suggest something else.

This particular Model S was caught on video sporting Velodyne’s roof-mounted LIDAR sensor near Tesla’s HQ in Palo Alto, California.

It’s clear from the video that the car is wearing a manufacturer’s plate, so it’s safe to assume that this is either Tesla’s car, or a car lent to a supplier for testing purposes.

Elon Musk insists radar is the way to go and with Autopilot 8.0 the radar-based system has been upgraded significantly for improved detection, but perhaps in the far off future LIDAR will make its way into Tesla’s lineup?

Via Teslarati

Tags: , , ,

23 responses to "Tesla Model S With Manufacturer’s Plate Caught Testing Velodyne Lidar System – Video"

  1. Kdawg says:

    Am I the only one thinking, “R2D2, are you OK back there?”

    https://youtu.be/BYUqXBYA7uM

  2. Josh Bryant says:

    Tesla using a LIDAR as a research sensor does not mean it will become a sensor for production vehicles in the future.

    I have a strong feeling Tesla is using the LIDAR point cloud data as validation data to improve the software algorithms that create point clouds from the radar signals.

    A secondary sensing system is always needed for validation.

    1. RM says:

      Bingo. It’s very likely for testing validation and/or comparison purposes.

    2. Kdawg says:

      For a production version, they would need to work on packaging. And I don’t know about the robustness of a rotating mirror. I like the solid state LIDAR research that is going on.

      1. TomArt says:

        So would Google, and everyone else that wants to use LIDAR. I don’t understand why the test mules continue to be tested like this when there is no way that the final product could look anything remotely like that – the geometry will be different and everything else!

  3. scottf200 says:

    Likely testing Lidar to compare it to their new radar system. Seems like another way to fine tune their radar system.

  4. speculawyer says:

    If Lidar provides better information, then adopt it.

    1. philip d says:

      It never will in even moderate rain or fog.

      1. Pushmi-Pullyu says:

        If the right wavelength(s) of light/radiation is chosen for lidar scanning, this isn’t true. Near-infrared-band lidar penetrates fog (and smoke) just fine.

        As far as rain, that interferes with radar as well as lidar… and the “mark one eyeball”, too. I’d like to see some evidence of how radar performs in rain vs. lidar. From my rather brief Google survey of the subject, I get the impression that the performance of lidar at least isn’t noticeably worse than radar.

        Some details of testing multiple commercially produced lidar scanner systems in “medium rain” here:

        http://robotsforroboticists.com/lidar-review-comparisons/

  5. DJ says:

    Well ya, radar is the way to go for Autopilot 8.0. Lidar on the other hand is the way to go for 9.0+ 😉

  6. gears says:

    How much does a Lidar system cost? My guess is a lot more, but might pay off for level 4 fleets

    1. Pushmi-Pullyu says:

      I don’t have figures on exact costs, but those “in the know” say lidar is considerably more expensive. Presumably that is why Tesla is resisting a move to lidar, despite the fact that Google’s more advanced self-driving cars use lidar… as does Lyft.

      Fortunately, development work is already in progress on solid state lidars, which should bring the price down considerably.

      One thing to consider is that a single lidar scanner in an elevated position above the car’s roof — which is what you see in the photo accompanying the article above — can scan in all directions, so only one is needed. Contrariwise, Tesla uses multiple radar sensors, and will be adding more to newer cars because the coverage of the current radars is not adequate.

      1. TomArt says:

        LIDARs will have to do the same thing. I refuse to believe that automakers are going to make production vehicles with a big, goofy &%(* LIDAR rig on the roof!

  7. ModernMarvelFan says:

    Is it me or that Model S is so close to the rear bump of that Ford Focus?

    Either way, I am glad that they are working on improvement.

    1. Bloggin says:

      It’s not Tesla thats doing Lidar testing, it’s Ford. Notice the Ford Escape and Ford Focus stacked very close to one another.

      Ford owns at least one Tesla Model S and Model X, and the vehicle would have manufacturer plates.

      From the video it looks like they are testing close proximity vehicles and the Lidar ability to identify the vehicles as separate units, as opposed to one big vehicle in the real world.

      If you think about it, the Model S makes a perfect shell on battery pack to test the technology. Ford has been very quiet about their upcoming 200+ mile EV, and it would be so easy to use the Model S as a mule. Especially since I suspected all along was that Ford and Tesla were quietly partnering on battery packs and other EV technologies.

      Remember, Tesla is coming with the Model 3, while Ford is coming with the Model E. There is a strong indication that these vehicles will share similar technology components.

      1. Pushmi-Pullyu says:

        Interesting speculation, but that’s all it is — speculation.

      2. TomArt says:

        That’s news to me…

  8. LS says:

    Definitely a Tesla Motors car. This license plate has been spotted on Tesla vehicules for years.
    http://www.motorauthority.com/news/1062269_2012-tesla-model-s-prototype-spy-shots-exclusive#image=100354267

    So probably benchmarking against Lidar.

    1. TomArt says:

      nice catch!

  9. Pushmi-Pullyu says:

    It’s good that Tesla is, apparently, at least testing out a lidar scanner system. As I’ve posted several times, I hope Tesla will move to a lidar system.

    The lidar scanner shown in the photo is properly placed, elevated above the top of the roof. Geometrically, this is a much better placement than the current radar scanners, which are (if my understanding is correct) mounted not much above bumper level. The roof mount lets a single scanner scan in all directions, and also gives a better “view” of the road. It’s the same reason that short women often prefer to drive an SUV or minivan; because the higher seating position puts their eyes higher, giving them a better view of the road. It’s simple geometry.

    But just because Tesla is testing a lidar scanning system doesn’t mean they have any real plans to change to that system. I wish I could think so, but this isn’t sufficient evidence to reach that conclusion. Tesla could be using this merely for comparison or verification purposes.

    1. TomArt says:

      I don’t know why everyone keeps testing LIDAR that way – I cannot imagine that any automaker would have a huge bulge on the roof for the LIDAR! Any sensor technology used in production vehicles would have to be body-level, not on top of the roof.

  10. Christian says:

    Music video “filmed” entirly with LIDAR:

    (the distortion effect is done with water)