If every car were driven by a computer, rather than a human driver, it is possible that a relatively simple set of rules could control vehicle behavior. Human drivers do not have the luxury of assuming that pedestrians, other motorists and conditions will behave in predictable ways. Google's testing of driverless vehicles has led to some interesting insights about what it takes to get around safely and efficiently on today's roadways.
For safety reasons, the current crop of self-driving cars being tested by Google come to a stop when confronted with conditions they are not programmed for. The unfamiliar brings the vehicles to a halt and forces human drivers to take over. It is a situation similar to that faced by inexperienced human drivers. The difference is that self-driving cars prioritize avoiding an accident over any other consideration. Young drivers face different pressures that may lead them to harm when they don't know what to do.
Google's testing has yielded other lessons about getting around. Several changes have been made to make the driver-less cars less passive. When driving in traffic, the vehicles no longer maintain the recommended distance between vehicles. As human drivers know, leaving a larger gap simply invites more aggressive drivers to cut you off, increasing the chances of an accident. Four-way stops have also posed a challenge, as human drivers are inclined to go out of turn if they think they can get away with it. The self-driving cars show that aggressive drivers force everyone into a more aggressive posture, simply to get around.
While there are obvious advantages to self-driving vehicles, it is not at all clear that they will ever completely replace human drivers. As long as people are responsible for making decisions on the roads, every driver must take those decisions into account in making the safest driving decisions.
Source: Top Tech News, "Google's Self-Driving Cars Encounter the Bizarre," by Steve Johnson, 17 November 2014