TMC is an independent, primarily volunteer organization that relies on ad revenue to cover its operating costs. Please consider whitelisting TMC on your ad blocker and becoming a Supporting Member. For more info: Support TMC
Start a Discussionhttps://teslamotorsclub.com/tmc/tags/

New Self Driving Demonstration Video (Is Tesla Ahead of Schedule?)

Discussion in 'Autonomous Vehicles' started by AssortedBread, Nov 18, 2016.

?

Is Tesla ahead of schedule? (on Autonomous cars)

  1. Yes

    53.6%
  2. No

    46.4%
  1. AssortedBread

    AssortedBread Member

    Joined:
    Oct 10, 2016
    Messages:
    203
    Location:
    Philadelphia, Main Line
    • Like x 2
  2. JeffK

    JeffK Well-Known Member

    Joined:
    Apr 27, 2016
    Messages:
    5,368
    Location:
    Indianapolis
    Love the music selection!
     
    • Like x 1
  3. Ulmo

    Ulmo Active Member

    Joined:
    Jan 19, 2016
    Messages:
    2,507
    Location:
    Vienna Woods, Aptos, California
    #3 Ulmo, Nov 19, 2016
    Last edited: Nov 19, 2016
    The car was unsure of itself quite a few times, stopping in illegal and extremely dangerous ways.

    I'm not seeing facial expression analysis to increase probability accuracies for questions like do they see you and do they see other vehicles to which they may or may not appropriately react, and how much attention is on new ideas like turning right or left.

    It was sped up so you can't see blinkers or other key signals.

    I can't tell from this display if they calculate speed, attitude, driver function of each car, to predict their possible motivations.

    I can't tell if they are willing to cross a double yellow line to avoid running over people. Do they cause stopping accidents maybe killing pedestrians in order to not cross double yellows?

    I can't tell if it honks at dangerous situations.

    I can't tell if it quickly starts to slow down slowly, watching rear cars to see if they react, and continue to monitor the rear cars progress at stopping while increasing stopping speed, until safely and comfortably stopped.

    I don't see multilane analysis, lane path probabilities, and other things about driving.


    But, it's looking better than the original video.

    The less they hand code and the more they generalize rules, the better: never sudden changes unless emergency; try to figure out what will happen, and plot course to probable destinations (skip stop at nearby store if rush hour traffic increasing; etc.) through this without hitting anything bad; attempt to follow laws. General rules like that would be better than hand coding each little thing. Pro race drivers and truck drivers could test its knowledge to make certain it is thinking correctly. Throw in there also chofeurs(sp) and extreme driving teachers to round it out.
     
    • Like x 1
  4. 3Victoria

    3Victoria Active Member

    Joined:
    May 8, 2016
    Messages:
    1,676
    Location:
    Victoria, British Columbia
    I love naysayers (not) with their continuous real time 'they won't be able to' comments, all the while abandoning their previous "can't"s as they are proved wrong without a nary apology. Tesla's products are in continuous development. AP does not need to be perfect, only demonstratively better than some level of present driving. Some say "this will never work in X situation" ... maybe they are right, maybe AP will not be offered in Mombay, say. That is not a fail.
     
    • Like x 2
    • Helpful x 1
  5. animorph

    animorph Member

    Joined:
    Apr 1, 2016
    Messages:
    514
    Location:
    Scottsdale, AZ
    Looks like a good step to me. It was hesitant about passing the two runners (walkers?) on the side of the road, and seemed to pause in a couple of strange places. But it handled some navigation and traffic.
     
  6. Canuck

    Canuck Well-Known Member

    Joined:
    Nov 30, 2013
    Messages:
    5,239
    Location:
    South Surrey, BC
    A computer program can be "unsure"? It looked to me like the program did what it was suppose to, and I'd much prefer to have AP2.0 driving other cars on the road than most humans. I also didn't see anything "extremely dangerous" but perhaps you can direct me to the times so I can take another look.
     
    • Like x 2
    • Helpful x 1
  7. hdgmedic

    hdgmedic Member

    Joined:
    Sep 29, 2015
    Messages:
    58
    Location:
    US
    So, this commentary is about what the video didn't show you ("I can't see...I can't tell....I'm not seeing...I don't see.... You can't see....") and not necessarily about what AP 2.0 can't do?
     
    • Like x 1
  8. JeffK

    JeffK Well-Known Member

    Joined:
    Apr 27, 2016
    Messages:
    5,368
    Location:
    Indianapolis
    #8 JeffK, Nov 19, 2016
    Last edited: Nov 19, 2016
    Every time is stopped it was for legit reasons (bicyclist coming up from behind and pedestrians potentially in the road.) except once at 1:32 (about 2:32 in paint it black version)... I cannot make heads or tails why it decided to make the turn then stop again.
     
    • Like x 1
  9. JeffK

    JeffK Well-Known Member

    Joined:
    Apr 27, 2016
    Messages:
    5,368
    Location:
    Indianapolis
    We don't yet know what it's incapable of...
     
  10. GoTslaGo

    GoTslaGo Learning Member

    Joined:
    Dec 25, 2015
    Messages:
    2,899
    Location:
    NV
  11. GoTslaGo

    GoTslaGo Learning Member

    Joined:
    Dec 25, 2015
    Messages:
    2,899
    Location:
    NV
    I think it stopped at that point because it rendered the double yellow line over to the left more than it actually was. This made the oncoming car seem like it was going to collide with the Tesla. Look at the middle frame on the right, you can see the red line overlay is to the left of the double yellow line. This makes the oncoming car (in blue) too close. At least that's what I can garner from the video.

    TeslaSelfDrive234.jpg

    Another thought is that the oncoming car was being designated with a both blue and a green box. Maybe that got the algorithm confused.
     
    • Informative x 1
  12. woof

    woof Model X 75D Blue, 6 seats

    Joined:
    Apr 30, 2009
    Messages:
    1,248
    Absolutely. It's all about probabilities, and if they get too low, or all the choices are equal, there has to be some default behavior or tie breakers.

    Interesting anecdote: while teaching my teen driving yesterday, on a 40 mph two lane highway we come across a car parked halfway into the road ahead. There is also an approaching car in the other lane. My son slows down to a complete stop, unwilling to cross the double yellow line while a car is approaching. I was terrified we were going to be rear ended (we weren't). Afterward while talking this choice over with him, he said he wasn't confident enough to pass the parked car with the other car coming. So by his calculations the probability of not hitting a car were too low, and the right response was to stop until the probability increased and he could pass safely. Experience will increase his confidence and feel for the amount of space he needs to pass, along with the experience of getting honked at or hit from behind if he stops where others aren't expecting. Two competing probabilities shaping the driving behavior.
     
    • Informative x 1
    • Like x 1
  13. Canuck

    Canuck Well-Known Member

    Joined:
    Nov 30, 2013
    Messages:
    5,239
    Location:
    South Surrey, BC
    You must have a different definition of "unsure" than the dictionary which is "not fixed or certain". If the computer program works on "probabilities" and has "default behavior or tie breakers", as you say, then you're making my point that the program is acting in a fixed or certain way, which is the way it was programmed to act. As such, it cannot be unsure.
     
  14. 3Victoria

    3Victoria Active Member

    Joined:
    May 8, 2016
    Messages:
    1,676
    Location:
    Victoria, British Columbia
    It could choice from equally weighted option using random choice o_O
     
  15. JeffK

    JeffK Well-Known Member

    Joined:
    Apr 27, 2016
    Messages:
    5,368
    Location:
    Indianapolis
    It won't, they've expressed that they want perfectly predictable and repeatable behavior. If all variables are the same then the outcome will be the same.
     

Share This Page