We do not know how Arthur C Clarke would have seen it but the death of a woman after being hit by a driverless Uber vehicle in Arizona in the US has certainly set the alarm bell ringing. The robot has claimed the first human life.
The problem we have in our hand is still a futuristic one. It scares us when we visualise millions of robots in the form of driverless cars move around on the streets and kill people made of flesh and blood at regular intervals. How do we stop it?
Uber Technologies Inc decided to suspend the road tests of its automated cars after the tragedy but the incident has raised a serious question over the risk that comes with automated cars. Are those speaking in favour of driverless cars doing so with full honesty and awareness?
According to David Fickling of Bloomberg Gadfly, other transportation technologies tell us that automated cars should become fine eventually but at the same time, the journey towards that safe driving would be marked with fatalities. Even in case of air transportation, crashes continued well after the Second World War, reiterating the fact that transportation safety comes with a cost as a natural law.
What should we learn from the Arizona tragedy about the future of automated driving on the roads?
First and foremost, one needs to give up the excitement for a revolution in favour of a more realistic evaluation of the driverless driving. Fickling says that just like conducting robotic surgery in a controlled environment, the safety benefits of driverless vehicle could be more than gained from better enforcement of speed limits or wearing seat belts. He says it would also be wise to learn that the initial benefits would be modest.
Secondly, it would be wise to take careful strides while making claims about transportation safety. Though the toll from accidents caused by autonomous cars is small at the moment, one should also keep in mind the fact that these cars are still in their early days. Fickling points out that there are not adequate data to back the safety claims and we should be aware of that.
Fickling also suggests an interesting angle while talking about the lessons. According to him, the auto industry should see more sharing of information so as to reduce accidents, as it is practised in the aviation industry but things are less transparent in the former where companies are reluctant to exchange such data so as to gain an advantage over the rivals. The less the database is shared among the players, the less likely is the number of accidents from coming down.