I first was introduced to this idea on an episode of Scientific American Frontiers in 1997 called ‘Robots Alive’. At the time a university had developed a vehicle that could drive itself down a highway. It could not navigate an intersection, parking lots, or too much confusion, but it was a car that drove itself! I was so excited that I proclaimed that soon you would be able to put a map into the car and that it would go where you wanted it to go.
I did not know that the map technology would come first, in the form of GPS. I had heard of the satellites that could pinpoint your position, and how amazingly accurate they were, but it did not occur to me that they could be overlaid on maps. Honestly I don’t think that I knew then that it was a global system. Today this is so ubiquitous that I get frustrated every time I have to give directions to the hotel where I work.
Many years later, after I had a drivers license, and my own car I began to wonder why do not have these self-driving cars yet. I have done a lot of cross-country driving, which is the technology that I knew specifically had already existed for at least ten years at this point. So I did a little digging. I found out that the research being put into this at the university was shut down. I could hardly believe it. It was shut down because of liability concerns. The thinking was that if one of these vehicles was deemed at fault for an accident the developers and programmers would be the ones held liable.
This opens up a whole other can of worms. In this country we are obsessed with making sure that someone else is to blame for every injury, no matter how small or who is actually responsible. I am no fan of McDonalds, but I do not believe that it is their fault that someone burned themselves on coffee that is obviously hot. This trend of suing others for your own incompetence is at best, stupid, and at worst, detrimental to scientific, technological, economical and societal progression. I am still in shock that the courts of this country have set something this damaging as a legal precedent in this country. It has gotten so crazy that homeowners can be sued by a burglar for injuries caused by hazards created by a burglar himself, and there are tales of this actually succeeding, although I have not found the records to back it up.
In order to get around this ridiculousness of American culture, when Google decided to create their self-driving car they pre-empted this argument. They have already established that the driver of the vehicle is legally responsible for any problems. This works because the car’s control can be overridden easily by the driver at any time by simply taking control as you would normally drive the car.
I also believe that because driving is really a routine of wrote behaviors and following specific signs. This is something that computers are inherently better at than humans are. ‘If this, then that’ is the way that computers think. Humans get distracted, intoxicated, bored, or try to multitask. These are the most common causes of accidents. Even the times that humans think of as a judgment call; for instance making the light before it changes, can be calculated with precision by a computer that we could never match. Driving is a computer-friendly activity. With a properly programmed design, I believe that accidents would largely become a thing of the past. On top of that other groups are working on technology that allows the vehicles to communicate directly with one another. This means that they do not need to anticipate each other’s actions or respond to last minute decisions. The main challenge at this point is accounting for human unpredictability.
When I learned about the Google self-driving car, I was very excited, and I still am. These cars, when I first looked several years ago had already logged over 100,000 miles on the open roads of California. They have all of the advantages of a computer on the road, and are still able to share the roads with human drivers and even avoid pedestrians and pets. In all that time, the only accident that the cars had was a fender-fender in a parking lot while driving under human control. With many more miles under their belts; nearly 1.7 million, they had only 6 accidents. None were caused by the autonomous vehicles (source).
There are other prototypes that would require all vehicles to be replaced with self-driving simultaneously. This is a completely unfeasible model to implement. Parts of the technology may be able to be adopted by others; like the inter-vehicle communications, but I cannot see all Americans giving up driving at the same time.
I am very excited to see these vehicles out there, making our roads safer. I have already seen cars that beep when the driver drifts out of their lane, a technology directly from the show that grabbed my attention so many years ago. There are cars that parallel park themselves, have blind spot monitoring, cruise control that adjusts as people ahead slow down. Tesla Motors actually has a car currently on the market that has an ‘autopilot’ feature that will “steer within a lane, change lanes with the simple tap of a turn signal, and manage speed by using active, traffic-aware cruise control“.
I am hopeful that these technologies will help us prevent accidents in the near future, and eventually be able to eliminate them. For now I am still behind the wheel. I do enjoy driving, but the more that I see of other drivers, the more concerned I am. I know that some drivers have much worse driving records than these cars do, and I believe that letting the cars replace these drivers would make us all safer. That said I totally understand being terrified of technology. I know how often computers malfunction. The automatic response to this type of eventuality must be considered in the design.