Making Self-Driving Cars Safer

We hear quite a bit about self-driving cars these days.  More and more cars on the road have at least some ability to do things on their own (steer, brake, or park) and some can do much more.

Inevitably, there have been a few accidents involving self-driving vehicles and they have attracted a great deal of attention.  There are about 6 million auto accidents in the U.S. each year, there are 3 million injuries as a result, and accidents kill roughly 35,000 people a year.  And yet, people are very worried about the handful of incidents involving self-driving cars.

The truth is that whenever a new activity results in people getting killed or injured, society questions whether we should be doing it.  This has occurred with every new mode of transportation from trains to jet planes.  What people often don’t consider is the safety of the new technology when compared with that of what it is replacing.

In nearly every case, collisions involving self-driving vehicles have been caused by human error, just as the great majority of all auto accidents are caused by human error.  In principle, self-driving cars should be much safer than people-driven cars since self-driving cars don’t text, get sleepy, or drink too much.

That being said, developers of self-driving technology need to work on giving the vehicles the ability to deal with unplanned situations.  Even if the cars are capable of creating an accurate representation of the space around them – which is still something of a challenge- they also need to be able to anticipate or at least be on the lookout for human errors in the cars around them.  It is necessary if self-driving cars and people are to coexist safely on the roads.

**********

Web Links

Redefining ‘safety’ for self-driving cars

Photo, posted May 28, 2014, courtesy of Flickr.

‘Making Self-Driving Cars Safer’ from Earth Wise is a production of WAMC Northeast Public Radio.