Announcement

Collapse
No announcement yet.

Drifting DeLorean

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Drifting DeLorean

    Mechanical engineers having fun: https://engineering.stanford.edu/mag...-driver-safety
    Allan Ostling

    Phoenix, Arizona

  • #2
    Not much of a loss to the automotive world considering a DMC-12 is worth about $22k.

    Be interesting to see what they did with the twisting X frame and where they mounted the roll bar.

    Most interesting is seeing the DMC sitting on an in ground 2 post lift in the land of environazis.

    Comment


    • #3
      I am less impressed with this drifting DeLorean than I was initially. The bot might be better at this partricular task than a human driver, but it is a task well-defined and marked with road cones. I'll change my mind when an autopilot can ride a trials bike like Malcomb Smith.
      Allan Ostling

      Phoenix, Arizona

      Comment


      • #4
        Autonomous robot-controlled car drifting through a parking lot and the thread didnt even make it one response before Goodwins Law comes into effect...

        Why cant people just be happy that we have robot cars driving themselves better than 90% of humans can manage

        Comment


        • #5
          Originally posted by aostling View Post
          I'll change my mind when an autopilot can ride a trials bike like Malcomb Smith.
          It's only a matter of time. Technology is not the best solution to every problem, but machines can do many things better than us. They don't get distracted, they can respond faster, see in the dark, communicate with each other at high speeds, etc. It's a matter of putting all the technologies together correctly. Yes, the machines get it wrong sometimes, but only because humans make mistakes designing, building and maintaining them. Machines like this are the culmination of million's of years of experience, research and innovation by humans. We can't evolve as quickly as we can improve the machines, so the machines will eventually get better than us at most things.

          There are downsides of course. Serious ethical questions about how far we should go. Unfortunately those questions will probably be brushed aside because the "other guys" will do it if we don't.


          Originally posted by epicfail48 View Post
          Why cant people just be happy that we have robot cars driving themselves better than 90% of humans can manage
          Because too many people here think they are in the top 10%

          Comment


          • #6
            Originally posted by epicfail48 View Post
            ........

            Why cant people just be happy that we have robot cars driving themselves better than 90% of humans can manage
            Perhaps once the robots ARE in that level of skill, people may be.

            The robots seem to be still coming up a bit short.
            1601

            Keep eye on ball.
            Hashim Khan

            Comment


            • #7
              Originally posted by aostling View Post
              I am less impressed with this drifting DeLorean than I was initially. The bot might be better at this partricular task than a human driver, but it is a task well-defined and marked with road cones. I'll change my mind when an autopilot can ride a trials bike like Malcomb Smith.
              Wouldn't it be just as impressive to see a robot bike that can do trials as well as YOU???? After all the first goal is to do as well as an average person on a good day. Then you move on to advanced things. When I watched the video and read the article, my takeaway was that the objective of drifting was to learn more about how to handle such maneuvers in an emergency, such as when hydroplaning or after the impact of an accident.

              An old adage: When given a dog who can talk, it's foolish to criticism his grammar.

              Dan
              Last edited by danlb; 02-01-2020, 12:36 PM. Reason: Original version sounded more confrontational than intended. Sorry Aostling.
              At the end of the project, there is a profound difference between spare parts and extra parts.

              Location: SF East Bay.

              Comment


              • #8
                Originally posted by J Tiers View Post

                Perhaps once the robots ARE in that level of skill, people may be.

                The robots seem to be still coming up a bit short.
                You think 90% of people can drift through a course without taking out a line of cones or spinning out? Wish I lived in a city with drivers like that, round here a light drizzle is enough to make everybody forget how to go in a straight line

                Comment


                • #9
                  Originally posted by danlb View Post

                  ... If you watched the video and read the article, you'd see that the objective of drifting was to learn more about how to handle such maneuvers in an emergency, such as when hydroplaning or after the impact of an accident.
                  I did read the article, and I did not mean to denigrate the achievement of those Stanford students. I was a student in that ME Department too, fifty-eight years ago. In 1962 there was one grad student working on how to control a roving vehicle on the moon, remotely from earth. The three-second time delay for a signal to get to the moon and back was a real challenge for any human at the controls. I forget how this came out, but the state of robotics at that time was equally incapable of steering the rover at anything faster than a slow walk.

                  Allan Ostling

                  Phoenix, Arizona

                  Comment


                  • #10
                    Originally posted by epicfail48 View Post

                    You think 90% of people can drift through a course without taking out a line of cones or spinning out? Wish I lived in a city with drivers like that, round here a light drizzle is enough to make everybody forget how to go in a straight line
                    Drifting is not the issue. It is a "straw man" argument. At least 99.999% of drivers have never needed to do a drift, or a "J" turn, etc. No point in considering that.

                    It's little stuff like not seeing the person with a bicycle, and fatally running them over...... mistaking a semi for a bridge and trying to go under it, not being able to negotiate streets that are not well marked, etc, etc.

                    That's all stuff that has come up with present day AVs and actually made the news. There will be lots more that has not yet killed anybody, so it only made the reports done by the "check drivers" monitoring the vehicles.

                    It's not knowing HOW to do a drift, it's knowing WHEN do do any of the huge number of maneuvers that are possible in a car.
                    Last edited by J Tiers; 01-31-2020, 04:51 PM.
                    1601

                    Keep eye on ball.
                    Hashim Khan

                    Comment


                    • #11
                      Originally posted by aostling View Post
                      I am less impressed with this drifting DeLorean than I was initially. The bot might be better at this partricular task than a human driver, but it is a task well-defined and marked with road cones. I'll change my mind when an autopilot can ride a trials bike like Malcomb Smith.
                      Ditto - in fact they said they ran multiple times so all in all they probably just ran a "program" and figured where the car went and then set up all the cones to make it "look impressive" lol Crush it and make some stainless steel kitchen utensils...

                      Comment


                      • #12
                        Originally posted by J Tiers View Post

                        Perhaps once the robots ARE in that level of skill, people may be.

                        The robots seem to be still coming up a bit short.
                        Estimate for fully autonomous self-driving cars by one of the Murata managers working on ADAS and autonomous driving systems: "Maybe year 2050 if ever"
                        Something 30 years in future is same "we don't even have clue yet" (fusion power anyone?)

                        Simple parking lot (or average US city for that matter) is pretty trivial compared to medieval european cities with narrow cobblestone streets, buildings overhanging on the passageway and pedestrians&bicycles bustling around in every direction.

                        Click image for larger version

Name:	image.jpg
Views:	117
Size:	172.4 KB
ID:	1852544

                        Possible still that trucks/semis (or lorries for brits) could leap in front of the autonomous driving systems at some point.
                        Few grand to several LIDARS is lot smaller part of the total budget in those.

                        Comment


                        • #13
                          Self driving cars need clearly discernable roadway edge markings to work, here is an interesting video.
                          https://youtu.be/3Y67XKPmtY8

                          Comment


                          • #14
                            Originally posted by J Tiers View Post

                            Drifting is not the issue. It is a "straw man" argument. At least 99.999% of drivers have never needed to do a drift, or a "J" turn, etc. No point in considering that.

                            It's little stuff like not seeing the person with a bicycle, and fatally running them over...... mistaking a semi for a bridge and trying to go under it, not being able to negotiate streets that are not well marked, etc, etc.

                            That's all stuff that has come up with present day AVs and actually made the news. There will be lots more that has not yet killed anybody, so it only made the reports done by the "check drivers" monitoring the vehicles.

                            It's not knowing HOW to do a drift, it's knowing WHEN do do any of the huge number of maneuvers that are possible in a car.
                            You mean the little stuff that human drivers do every single day like run over bicyclists and pedestrians, cross the center line, run stop lights, etc.? A coworker of mine hit a bicyclist one morning after night shift. He came over a rise and the sun momentarily blinded him. He was doing the speed limit etc. and wasn't charged. Has to live with that of course. But now if a self driving car was involved who do think would get the blame. The car or the programmer of course because they are supposed to be perfect. We can except human failure but not failure of machines we build. This is really more of a social issue than a technology issue.
                            Tom - Spotsylvania, VA

                            Comment


                            • #15
                              Originally posted by flathead4 View Post

                              You mean the little stuff that human drivers do every single day like run over bicyclists and pedestrians, cross the center line, run stop lights, etc.? ....
                              Yes, if there is to be a point to the AVs, they should be better than human drivers. Otherwise we are setting an acceptable number of designed-in deaths.

                              Many or most of the existing traffic accidents are consequences of things people intentionally do.... driving drunk, speeding, texting etc. They are not actually due to deficiencies of human perception, etc.

                              Essentially ALL of the AV "accidents" have been programmed-in. They are due to perceptual failures, with a few possibly due to programmed-in bad choices of action.
                              1601

                              Keep eye on ball.
                              Hashim Khan

                              Comment

                              Working...
                              X