This recent statement from US Secretary of Transportation’s Ray LaHood is encouraging: "Whether we're talking about automated features in cars today or fully automated vehicles of the future, our top priority is to ensure these vehicles – and their occupants – are safe” The key will be the actual definition of “safe” that is adopted. Nothing is perfectly safe, and as we all know the existing surface transportation systems is far from safe. Mechanically, it does just fine. The safety issues flow from human frailty, and we would not accept the safety shortcomings were it not for the phenomenal mobility that our system delivers. Hopefully, the Federal policy will allow and encourage the substitution of smart driving technologies that are better and more reliable even if they are not perfect. We will all be safer with policies that allow the marketplace to offer these technologies to enhance our own capabilities or to replace them in places and at times when we have chosen to accept the responsibility. Consumers have already begun to recognize this by purchasing cars with anti-lock brakes (we allow the automated system to apply the proper brake function so as to do a better job of slowing down than we would have done) and by accepting airbags that deploy automatically in a crash (we are simply not good at doing that task when it needs to be done). Similarly, if we are about to crash, a smart automated system should take over and not require us to “remain vigilant” or to handle a situation that can be better handled by a system that is dedicated to the task. With dedicated automated technologies that can handle lane keeping, car following, and emergence response we would, in many if not most driving situations, be more safe than we are now when we break the rules and use our phones or become distracted by the scenery. Policy makers should use the “carrot” approach of encouraging the automobile industry to offer automated technologies that unshackle the driver from the routinized tasks of driving. Carrots will do more to improve safety than the current “sticks” of minimally enforceable anti-distraction laws and mandates for anti-communication gizmos that are easily disabled.
NHTSA Preliminary Statement of Policy Re: Automated Vehicles http://orfe.princeton.edu/~alaink/SmartDrivingCars/Automated_Vehicles_Policy.pdf contains the details of this preliminary policy. I highly recommend that you read it. My interpretation is as follows:
1. This is NHTSA’s first effort to play catch-up in what it has finally recognized as a crash avoidance oversight responsibility. This is a major extension of its traditional focus on crash mitigation. Up to now, for crash avoidance, NHTSA has largely concentrated on modifying driver behavior rather than on the testing and certification of technologies that compensate for human frailties. The notable exceptions are anti-lock brakes and active vehicle stability control. Some in the automobile industry as well as others such as Google are far ahead of NHTSA in the development of technologies that override driver actions or inactions that would otherwise have resulted in a crash.
I have 6 more points at: http://orfe.princeton.edu/~alaink/SmartDrivingCars/CommentOnNHTSA_PrelimStatement.pdf