The Globe and Mail reports on a recent survey of Canadians concerning their views on self-driving cars, apparently meaning cars with advanced driving assistance technologies such as the Autopilot mode in a Tesla.
If you have read my earlier post about new safety features planned for Volvo busses, then the results of the survey may not surprise you. Basically, many Canadians would take advantage of the features to do things they see as more productive than simply driving:
The survey of more than 2,600 Canadians found that 9 per cent of those surveyed said they would drink and drive; 10 per cent would sleep or nap; and 17 per cent would do something unrelated to driving. ... Another 35 per cent of those surveyed said that if they were late, they would disengage self-driving technology in order to drive faster and 13 per cent would take over so they could run red lights.
Clearly, these attitudes are bad news for road safety, just the thing these features are supposed to promote!
The Traffic Injury Research Foundation, sponsors of the survey, feel that the problem is that Canadians overestimate what services such as the Tesla Autopilot can do. Thus, educating them on this matter would improve the situation.
No doubt Canadians do overestimate the capabilities of current "self-driving" cars. Yet, it is not clear that education can repair the situation.
Consider the case of distracted driving: People are generally aware that it is dangerous to text and drive yet they do it anyway. There are many reasons. Distractions from smart phones are quite compelling. Mostly, people get away with doing it, and the risks of being penalized are small. Drivers feel that they are better-than-average multi-taskers. They feel entitled to use the roadways as they see fit.
These matters are well examined in a recent panel discussion on TVO's The Agenda:
Interestingly, design scholar Donald Norman has recently recommended partial automation of driving as a response to distraction, a reversal of his earlier views. He argues that the risks posed by distracted driving are increasing so much that the risks of inattention due to over-trust in automation pale in comparison.
He could be right. However, the new survey suggests that a kind of feedback could develop. That is, advanced safety features may license riskier driving, which would call for more aggressive safety features, etc. Can the developers of safety technology win this race?