Back in 2018 Uber caught some well deserved for flak an incident that occurred in Arizona involving one of their experimental self-driving cars. I'm sure anyone who's ever seen a sci-fi movie can guess what happened. The car evidently lost control and killed a pedestrian. This is far from the first time a self driving vehicle test has gone wrong but this does appear to be the first incident of a death resulting directly from automated driving technology.
 |
| Uber Self Driving SUV Before Crash |
 |
| After Crash |
The fatal accident took place on the night of March 18th, 2018 in Tempe, Arizona. 49-year old Elaine Herzberg was struck and killed while cycling across the road. Was Uber held responsible for the accident? Of course not. The autonomous car had a back-up driver behind the wheel in case of emergency situations. According to the company the auto braking system was not enabled for the computer driven mode. This decision was made in an attempt to fast track Uber's self driving taxi service by giving the effect of a smoother ride, previously the program made the car stop frequently making for a bumpy journey. This meant that true emergency breaking was left to the back-up driver who was not aware of their surroundings at the time. It was discovered that the car registered an "unknown object" which it shortly after recognized as a bike six seconds before collision. This system did not alert the driver, because the "shape" it had initially detected was classified as "benign", who looked up one second before impact and attempted to swerve to miss the cyclist. In other words the braking program had been redesigned to take risks and chances. The driver had been streaming popular TV hit
TheVoice before the crash which didn't help their chances of getting away Scot-free . Due to possible negligence on their part they were being faced with possible manslaughter charges.
In the end Uber got away with nothing more than a settlement with the victims family members. But as more accidents involving self driving cars occur the research seems to keep moving forward uninterrupted. This brings forth classic ethic dilemmas, should a machine be able to make potential life and death decisions. If a school bus is out of control and hurtling down the wrong lane can a machine possibly sacrifice the life of it's driver by veering off the road to save the live's of the children on board? Science fiction writer Issac Asimov has prompted questions like this for many years with his Three Laws of Robotics.
Something else that may help outline possible problems with self driving vehicles is the
Trolley Problem. In short the Trolley Problem is just an example used in ethics studies to show the logical, sentimental and/or duplicitous nature of human beings. Part of being a human is free will. If you want to run family of three off the road and kill them rather than chucking yourself off a cliff to let them survive you have that choice. Obviously the opposite could be true as well, a human could use their free will to save lives. But in the end this tech essentially takes the choice out of people's hands. This is a argument that is liable to go on for the rest of human existence, the bare bones question is, is it ethical to let machines decide who lives or dies based on programming?
New York Times Report On Uber Crash
NPR's Take
Streetblog.USA Negative Take
No comments:
Post a Comment