SoylentNews

Uber Pulls Self-Driving Cars After First Fatal Crash of Autonomous Vehicle

posted by Fnord666 on Monday March 19 2018, @10:31PM   
from the bound-to-happen dept.

vux984 writes:

A self-driving Uber SUV struck and killed a pedestrian in Tempe, Arizona. It was in autonomous mode at the time of the collision, with a vehicle operator behind the wheel. Uber has suspended testing of its self-driving cars.

http://money.cnn.com/2018/03/19/technology/uber-autonomous-car-fatal-crash/index.html
https://www.nytimes.com/2018/03/19/technology/uber-driverless-fatality.html
https://finance.yahoo.com/news/self-driving-uber-kills-arizona-171055918.html
https://www.npr.org/sections/thetwo-way/2018/03/19/594950197/uber-suspends-self-driving-tests-after-pedestrian-is-killed-in-arizona
https://www.wsj.com/articles/uber-suspends-driverless-car-program-after-pedestrian-is-struck-and-killed-1521480386
https://www.theverge.com/2018/3/19/17139518/uber-self-driving-car-fatal-crash-tempe-arizona
https://techcrunch.com/2018/03/19/uber-self-driving-test-car-involved-in-accident-resulting-in-pedestrian-death/

I couldn't find any good analysis of the liability situation here.

---

Original Submission

• I don't agree with you very much at all (Score: 0) by Anonymous Coward on Tuesday March 20 2018, @08:03PM

  by Anonymous Coward on Tuesday March 20 2018, @08:03PM (#655600)

  That's unrealistic given the way gasoline automobiles work.

  I don't agree at all with this statement.

  It doesn't take much external force to alter the car's velocity a small amount.

  And if the control loop on those cars is under 1khz I'd be surprised - that means with in 1/500th of a second corrections can be applied to any external disturbance. Running at 1khz is trivial for the velocity control loop if they wanted to.

  The nature of the drive train means over & under corrections are inevitable.

  True but that is actually the nature of any control system. They always go over and under their set point and perform constant adjustments when that happens. They effectively always have error that is always being corrected with an average that is the correct value. This is a standard problem solved all the time.

  A slight rise or dip in the terrain or a gust of wind would be plenty to change the car's velocity by a few MPH.

  The sensors are accurate enough that the deceleration can be measured as soon as the car starts traveling up hill at a fraction of a degree. With in a tiny fraction of a second the throttle will be adjusted and another reading from the sensors is taken.

  The physical system that the computer is controlling (engine & brakes) isn't accurate enough to avoid inevitable overshoot when the computer applies control.

  Control systems always overshoot, this is not realistic to complain about. Perhaps you could say the performance is too high and the car will overshoot too far and not maintain the speed requirements but that's a problem the control loop tuning solves.

  The only way you'd get a control system to keep the car exactly at a given speed would be to have it constantly alternate gas & brake to speed up and slow down.

  Is that how you drive? Jesus christ - you are always on the throttle or brakes? Do you even have a license or drive a car?

  A car constantly braking & accelerating would appear erratic to other drivers, probably resulting in them speeding around it to "get away from" the erratic driver.

  Dude the brakes and the gas pedal are not binary.

  Until/unless 100% of cars become fully autonomous, self-driving cars need to behave in a fashion that's similar to what human drivers expect of other human drivers. Drifting a few MPH over or under the posted limit is normal, and thus it's actually the safest way for an autonomous car to behave. The reactions of human drivers around it are the most likely source of accidents.

  Well moving above the speed limit because it's safer in traffic isn't really a thing to worry about in the middle of the night, now is it?

  Parent