Group A - autopilot drives
Group B - participant drives
First few sessions, nothing happens. Then a dog, human, car... appears - situations that self-driven cars have been known to malfunction.
Aim of the study is to gauge reaction times between these two groups.
It will help us know what to do with bored - consequently distracted drivers caught in accidents.
We can have a realistic third group. An autopilot group whose initial sessions show the car successfully handling two or more near-misses.
But the massive number of fatalities that occur every day in cars are rarely or not even mentioned by local news. We need fewer automation features and more assistive features like braking.
More hardware and more lines of code written by fallible humans isn't going to take human drivers out of the equation in my lifetime.