In the article below the author mentions instances that pose possible problems for future self driving vehicles in America. Can these vehicles be programmed to include the anxiety perimeter of a passenger in addition to the conventional road safety data. The author mentions a trip to another city and being dropped off at night by a cab driver one block from his destination next to a wall with graffiti on it. He was freaked out over the walk to his hotel in a part of town he did not know.
An entrepreneur I knew from Montreal visited an international trade fair in Miami. He was told by local residents attending the fair to watch out driving through sketchy neighbourhoods. His hotel was on the other side of one of these as it happened. Late that night he stopped his rental at several red lights in the afor mentioned neighbourhood as he made his way to his hotel. At the last one a pan handler approached his car. The driver checked to see that all his windows were up and doors locked. The homeless man stood waiting on the drivers side. The driver had been told to floor it through red lights if necessary in exactlly this type of situation and so he did leaving the man standing hand outstretched. He told me that he was told street people often work in teams. The leader will try to get access into the car by opening a door or by reaching into an open window. If he succeeds at this his pals come out of hiding to help. You can then expect to be robbed and have your vehicle car-jacked leaving you stranded on the street with nothing but the clothes on your back. The question is can self driving vehicles being programmed to respond like this driver did based on fears real and or imagined.
Ah, the little things. We rarely think thru what we want to do. The front end always is exciting. Back end not so much. Same can be said about combustion cars. The power to move looked good. Took a long while to figure out that the shit left over post move was a problem.