Tesla training autopilot to recognize lane splitting motorcycles

Discussion in 'Electric Motorcycles' started by voltsxamps, Dec 20, 2018.

  1. voltsxamps

    voltsxamps Advolturer

    Mar 20, 2016
    +positively grounded-
    The following article is based on a Tesla Model 3 owner's youtube post of updated autopilot software being able to recognize motorcycles lane splitting. It was also mentioned by the youtuber that utilizing audio sensors to better detect motorcycles may help, although, there could be some erroneous audio readings or the lack thereof as electric motorcycles become more common, especially in states like California, where not only lane splitting is legal, but the most likely place you'll find vehicles with autopilot as well as electric motorcycles:

    "Tesla Autopilot is increasingly starting to be able to detect and understand some “corner cases” that it encounters on the road and now we get a good look at how it detects an interesting situation: lane-splitting motorcycles.

    Lane-splitting motorcycles can be hard to detect even for experienced human drivers.

    Now Tesla is trying to train a neural net to detect them and Model 3 owner Scott Kubo did some tests to how they are doing right now.

    With the launch of its Version 9 software update, Tesla deployed a much more advanced Autopilot neural net and it started to more accurately render the vehicle’s surroundings on the instrument cluster.

    Tesla vehicles with the latest software and the Autopilot 2.0+ hardware suite can render cars, people, motorcycles, and trucks all around the vehicle:


    It makes drivers better understand what Autopilot can detect around your vehicle, which led Scott to put it to the test with lane-splitting motorcycles.

    With the Tesla software version 9 (2018.42.3), he filmed several motorcycles pass by his Model 3 to document if and how the Autopilot detects them.

    It resulted in a pretty interesting video with mixed results:

    (^ interestingly, these happen to be the exact same SoCal freeway locations that I travel on)

    Electrek’s Take
    While it’s still far from perfect, I think this is quite impressive and shows promise.

    On a few occasions, you can actually see the vehicle coming behind on the display. Sometimes it’s mistaking it for a car and even putting it in the next lane, but it’s likely something that they could fix by training the neural net to recognize the behavior of a lane-splitting motorcycle.

    Even if the cameras didn’t appear to detect the motorcycle, or at least didn’t render it on the screen, it looks like the ultrasonic sensors did, which would have probably resulted in a blind spot warning if he would have signaled to change lane.

    Scott also had a very interesting point about using audio to detect upcoming motorcycles. Humans do it, but I am not aware of an autonomous driving or driver assist system that makes use of audio.

    That could be interesting – though increasingly less relevant as motorcycles become electric."

    alternate article: https://www.rideapart.com/articles/298315/tesla-autopilot-detect-motorcycles/

    ^ interesting solution being proposed to use sound, though there are occasionally 4-cylinder imports with fart can exhausts that sort of also sound like a inline 4 motorbike, but to the author's point, it's unlikely that electric motorcycles would be audibly detected.

    I applaud any auto manufacturer, Tesla in this case, looking into identifying and avoiding motorcycles, including those that are lane splitting. As a lane splitting motorcyclist in California and as a Tesla Model S owner, I am interested in any tech that may help avoid collision and lessen the already inherent dangers of motorcycling. Especially as more vehicles become equipped with self drive capability and more drivers utilize it. This article addresses the very thing that goes on in my mind (and many other motorcyclists I've spoken to in CA) each time I lane split pass a Tesla or another AP equipped vehicle, wondering if 1) if that particular vehicle is on autopilot, and 2) if the autopilot recognizes that I am there.

    "On July 27, 2016, a motorcyclist in Norway was seriously injured when she was rear-ended by a Tesla model-S with its Autopilot engaged. The car went in for further testing. On December 7, 2017, California motorcyclist Oscar Nilsson was hurt in a collision with a self-driving Chevy Bolt"

    "don't ride significantly faster than the traffic you are passing. Not only will Autopilot be unable to detect you, neither will human drivers."

    ^ https://www.rideapart.com/articles/264772/opinion-on-autonomous-cars-motorcyclists-and-culpability/

    Perhaps sensors on electric motorcycles can be part of the bike's build to communicate with autonomous drive capable vehicles if necessary. As it stands, if a vehicle that is running semi or full autonomous drive capability on older AI hardware decides to change lanes, (and the driver isn't paying attention as they are supposed to) it could very well collide with a motorcycle. It's worth mentioning that only Tesla's with AP produced after Oct. 2016 have the capability to upgrade to this ability to recognize lane splitting motorcycles, and new computer processors due to come out next year will be up to ten times faster than the current ones, which may help.

    Interesting times we live in.
    RCmoto likes this.
  2. ctromley

    ctromley Long timer

    Oct 18, 2012
    I am tremendously grateful to Tesla and Elon Musk for putting EVs on the map. But it needs to be said that calling Tesla's driver aids "Autopilot", and pretending it is safe to use them as anything but driver aids is at least beyond idiotic, and at worst criminally negligent. Every time someone gets hurt or killed when an AV is involved is yet more proof that AVs are not yet ready for prime time. If you're thinking that the drivers are warned that they need to pay attention, understand the only way that makes sense is if you ignore the very natural tendency of us humans to rely on the convenience you're selling us. Any Human Factors engineer will call bullshit on that line of reasoning in a heartbeat.

    Properly functioning, truly autonomous and safe AVs are a SPECTACULARLY complex challenge. We are nowhere near achieving it yet. The hype is WAY beyond the reality, and it's literally killing people.

    Anyone operating any "autonomous" vehicle needs to assume it is not autonomous at all. Verify the information it provides, knowing it may be wrong or incomplete, and make your own decisions. You are the captain of your ship, and you are responsible for what happens as a result of its travel through space and time. Anyone telling you otherwise is lying.

    And be aware there will be plenty of other "captains" out there who are completely ignorant of these simple truths.
    MJSfoto1956 likes this.