At the Google headquarters in Mountain View yesterday, California governor Jerry Brown signed the bill, “SB-1298 Vehicles: autonomous vehicles: safety and performance requirements“, making California the third state in US taking steps to make driverless cars eventually legal on roads and highways (not just for testing purpose).
Technologically, Google will need to do a lot of testing to make sure its system works in many different basic conditions and to prove that it is “safe” in basic conditions which is a challenge in itself. The bigger problem, at least as this reporter sees it, is a legal one.
As Google and its co-founder Sergey Brin indicated, the hope is to allow the driverless cars to be used by blind people and people who are too intoxicated to drive. Imagine a Google driverless car involving in a car accident (especially one with fatality), who will then be held accountable? The blind driver? The software developer? The software project manager? Or the autonomous vehicle technology providers (like Google)?
Comparison has been made of Google driverless car with self-driving train and that we have those trains for some years now, isn’t it ok? Well, self-driving train (or centrally controlled driver-less train) like the SkyTrain in Vancouver runs on “grade-separated tracks, running mostly on elevated guideways” thus are not subject to the randomness of roads filled by other cars, random animals, and most important of all, often unpredictable people/pedestrians.
In a litigious place like California, United States, it is worth mentioning one of the two unresolved issues in the signed bill (as stated in the 08/24/12 Assembly Floor Analysis, emphasis added),
“First, automobile makers fear that when this technology is added on to the vehicles they have designed for use by human operators, they may be drawn into liability litigation after an accident caused by the failure of that after-market equipment or of its installation rather than due to any inherent safety problem with the vehicle itself. Although existing tort law and case law can be interpreted to absolve them of any liability in such instances, the automakers sought to have explicit language incorporated into the bill to make this clear. Ultimately, the contending parties were unable to agree on a formula that met one another’s needs.“
California legal experts may have better insight of why the above issue remains unresolved. To the average Joe and Jane citizens, a valid question is whether Google or other autonomous vehicle technology providers should be trusted to be 100% correct all the time with no driver to provide override in the future (when blind/intoxicated people may be the only passengers in the car)? Given the automobile makers are worried enough to try (but unsuccessfully) to shield themselves with explicit language in the bill, may be citizens should ask Google or the legislators what happen if a driverless car is involved in an accident with/without fatality?
Note: Currently, the California bill has a five million dollars insurance provision in this testing phase when there will ALWAYS be a driver on the driver seat to take over when things go wrong. What amount of “evidence of insurance, surety bond, or self-insurance” should be required when the system is truly driverless?!
P.S. this article is cross-posted by me at examiner.com