4 ms·

Not having LiDAR for anything over level 3 self-driving capabilities seems like a very bad idea...Computer vision right now just does not have the spatial aware

Not having LiDAR for anything over level 3 self-driving capabilities seems like a very bad idea...Computer vision right now just does not have the spatial awareness that you need for self driving capabilities. I wish Tesla would work on something similar to the Kinect v3 but for long range (above 20 m).

close04 8y ago
Elon insists being theoretically right is the same as being practically right. So for example he uses the theoretically perfectly functional normal vision that humans use even if the practical implementation of today simply cannot reach the same results.
- erikpukinskis 8y ago
  You think Elon Musk doesn't think there's any difference between theory and practice? What makes you think that? And what does it meant to be "practically right" about the future? I don't understand how you could be right in practice about something that hasn't happened yet.
leesec 8y ago
Having LiDAR in any car for self driving seems like a prohibitively expensive idea though. For instance: See how no company has released a commercial product with LiDAR. And I definitely think Computer Vision will get better before LiDAR gets cost competitive.
- Robotbeat 8y ago
  I agree. However, suppose the opposite: Suppose LiDAR gets cheap first. Like, I don't know, $1000 per unit and compact enough to not be a big burden on the rest of the vehicle. Tesla charges $5000 for the FSD add-on. In principle, they could easily afford to retrofit vehicles with the cheap LiDAR plus maintain any advantages the full computer vision system would add (such as identifying vehicle types to assist in predictive behavior modeling and avoidance strategies). So trying to go the full Computer Vision (with ancillary radar) route actually has pretty low opportunity cost for Tesla. Another thing: I wonder how much simply much-improved geolocation could help? That failure from last year of driving right into a (failed) collision attenuator (where an off-ramp split off) could've been avoided simply by having high confidence, half-meter-or-better geolocation (from GPS, cell, IMU, etc fusion) and good mapping. And with good mapping/geolocation combined with other sensors, the sensors could focus on identifying changes to the expected road condition, perhaps increasing their robustness. (and when networked, they could use sensing from other vehicles ahead of the current vehicle to assist in navigation as well)
  - leesec 8y ago
    AFAIK all the big companies working on this problem include high-definition mapping as a major component. The problem is gathering that data without having an actual LiDAR. So some companies like Waymo will literally just have cars with LiDARs drive all around every street and map it for them. This is great but means they're limited by when they travelled the road and conditions at the time. Ideally, you have a mapping system that is constantly updating (every time your fleet drives by something new like construction on the road it auto-updates the map). Comma.ai is working on a vision+GPS Kalman filter solution along these lines up to 10 cm accuracy. I'd guess Tesla is as well. And then yes, after that the vision would be primarily for localization.
    - Robotbeat 8y ago
      Indeed, as recent Tesla patent(s) show: https://electrek.co/2018/12/09/tesla-patent-technology-accurate-gps-positioning-for-self-driving-vehicles/ https://electrek.co/2018/12/09/tesla-patent-technology-accur...
- syntaxing 8y ago
  Totally agree. The race is a reputable solid state LiDAR which Waymo [1] and Velodyne announced a new one of recently [1]. But still, those are around 5K. Though I am really excited for Bajara's prism based scanner [3]. [1] https://techcrunch.com/2019/03/06/waymo-to-start-selling-standalone-lidar-sensors/ https://techcrunch.com/2019/03/06/waymo-to-start-selling-sta... [2] https://velodynelidar.com/velarray.html https://velodynelidar.com/velarray.html [3] https://www.baraja.com/ https://www.baraja.com/
alkonaut 8y ago
Regardless of whether cameras or lidar win out, just consider the fact that back up cameras already exist on tons of modern cars, and the solution to the environment issues of these cameras is one of: a) fold it out when you put the car in reverse, so it's always ice-free b) ignore this problem. It's covered in ice for months. This is a simple problem in comparison, and neither of the above solutions would work for a camera used for autonomous driving.
- ec109685 8y ago
  If the camera is broken, Tesla will obviously not enable auto driving. It has a steering wheel for a reason.
  - alkonaut 8y ago
    If it’s “broken” every morning for months in winter climates like the cameras I mention, then people there won’t pay a premium for the capability. And if it doesn’t keep itself ice free it will risk going from uncovered to covered in seconds. Cameras need to be ice/snow/rain proof in the sense that they stay clear of it while driving.
    - ec109685 8y ago
      What about a heated camera lens?

close04 8y ago

Elon insists being theoretically right is the same as being practically right. So for example he uses the theoretically perfectly functional normal vision that humans use even if the practical implementation of today simply cannot reach the same results.

erikpukinskis 8y ago
You think Elon Musk doesn't think there's any difference between theory and practice? What makes you think that? And what does it meant to be "practically right" about the future? I don't understand how you could be right in practice about something that hasn't happened yet.

leesec 8y ago

Having LiDAR in any car for self driving seems like a prohibitively expensive idea though. For instance: See how no company has released a commercial product with LiDAR. And I definitely think Computer Vision will get better before LiDAR gets cost competitive.

Robotbeat 8y ago
I agree. However, suppose the opposite: Suppose LiDAR gets cheap first. Like, I don't know, $1000 per unit and compact enough to not be a big burden on the rest of the vehicle. Tesla charges $5000 for the FSD add-on. In principle, they could easily afford to retrofit vehicles with the cheap LiDAR plus maintain any advantages the full computer vision system would add (such as identifying vehicle types to assist in predictive behavior modeling and avoidance strategies). So trying to go the full Computer Vision (with ancillary radar) route actually has pretty low opportunity cost for Tesla. Another thing: I wonder how much simply much-improved geolocation could help? That failure from last year of driving right into a (failed) collision attenuator (where an off-ramp split off) could've been avoided simply by having high confidence, half-meter-or-better geolocation (from GPS, cell, IMU, etc fusion) and good mapping. And with good mapping/geolocation combined with other sensors, the sensors could focus on identifying changes to the expected road condition, perhaps increasing their robustness. (and when networked, they could use sensing from other vehicles ahead of the current vehicle to assist in navigation as well)
- leesec 8y ago
  AFAIK all the big companies working on this problem include high-definition mapping as a major component. The problem is gathering that data without having an actual LiDAR. So some companies like Waymo will literally just have cars with LiDARs drive all around every street and map it for them. This is great but means they're limited by when they travelled the road and conditions at the time. Ideally, you have a mapping system that is constantly updating (every time your fleet drives by something new like construction on the road it auto-updates the map). Comma.ai is working on a vision+GPS Kalman filter solution along these lines up to 10 cm accuracy. I'd guess Tesla is as well. And then yes, after that the vision would be primarily for localization.
  - Robotbeat 8y ago
    Indeed, as recent Tesla patent(s) show: https://electrek.co/2018/12/09/tesla-patent-technology-accurate-gps-positioning-for-self-driving-vehicles/ https://electrek.co/2018/12/09/tesla-patent-technology-accur...
syntaxing 8y ago
Totally agree. The race is a reputable solid state LiDAR which Waymo [1] and Velodyne announced a new one of recently [1]. But still, those are around 5K. Though I am really excited for Bajara's prism based scanner [3]. [1] https://techcrunch.com/2019/03/06/waymo-to-start-selling-standalone-lidar-sensors/ https://techcrunch.com/2019/03/06/waymo-to-start-selling-sta... [2] https://velodynelidar.com/velarray.html https://velodynelidar.com/velarray.html [3] https://www.baraja.com/ https://www.baraja.com/

alkonaut 8y ago

Regardless of whether cameras or lidar win out, just consider the fact that back up cameras already exist on tons of modern cars, and the solution to the environment issues of these cameras is one of: a) fold it out when you put the car in reverse, so it's always ice-free b) ignore this problem. It's covered in ice for months. This is a simple problem in comparison, and neither of the above solutions would work for a camera used for autonomous driving.

ec109685 8y ago
If the camera is broken, Tesla will obviously not enable auto driving. It has a steering wheel for a reason.
- alkonaut 8y ago
  If it’s “broken” every morning for months in winter climates like the cameras I mention, then people there won’t pay a premium for the capability. And if it doesn’t keep itself ice free it will risk going from uncovered to covered in seconds. Cameras need to be ice/snow/rain proof in the sense that they stay clear of it while driving.
  - ec109685 8y ago
    What about a heated camera lens?