Welcome to the blog of Can Sar, a Stanford CS major. This blog is made up of my thoughts on Computer Science and the computer industry, as well as ever exciting tales from my life.
This is a response to Professor John McCarthy’s essay on Personal Flying Machines for my CS 22 N class, taught by him.
There are several important issues raised by the possibility of building Personal Flying Machines, and they can roughly be divided into 2 categories: Technological Challenges and Social Implications.
As stated in the essay, most humans cannot safely control airplanes or other flying machines. It will therefore be necessary to have some kind of automatic computer control. Each aircraft will have to be operated independently, because a central controlling system would have a number of problems.
First of all, security and fail-safety are major problems. Even though it would be possible to have backup systems in different locations in case one of them were to be damaged, this would not eliminate all problems. If the central control system had a security leak, the existence of backup systems would not be of any use, because they would only come into place. This can of course be made less likely, by having some of the backup systems monitor the course of each aircraft and take over in case one or more of the backup systems disagree with the main one, over how to control the object. With sufficient refinement this would be quiet secure. People might also have problems with flying in something that is remote controlled, but this feeling would probably go away, once people got used to it.
There is, however, a fundamental problem with such a central system. It would cost large amounts of money to construct, and would be hard to upgrade to scale to larger number of users. This would mean that the first users would have to suffer the huge costs of building this system, before their aircraft becomes really functional. It would also mean, that some independent organization would have to be set up and countries make arrangements on how their control systems would interact. Deciding what kind of organization should do this would be incredibly confusing, and ensuring security in all countries would be very difficult.
A distributed system is a much better alternative.
Each car would measure would measure the distance to surrounding objects, air movement, local weather data, and other information. This could be combined with some data like GPS and weather information, to give the aircrafts computer enough information to make decisions on how to fly. This would make the system quiet usable with only one such aircraft. The only thing that would be needed would be 2 landing platforms, or even just 1, if the aircraft would just be used for recreation and sight seeing, rather than travel. Of course adding more platforms would make the system more useful, but the cost of doing so is not nearly comparable to setting up a central system for controlling aircraft, before even a single one of them is in use.
I believe that making these aircraft maneuver automatically while in the air, will be possible. Current autopilots are very effective, and are able to control aircraft most of the time. Depending on the exact nature of the aircraft, piloting it automatically can become more difficult, which is the case with helicopters, but it is still easier than landing them. Until personal aircraft become a reality, this problem of finding computers smart enough to land must be solved. Wind forces are especially dangerous at that time, and must be carefully accounted for. Once this can be done with computers, personal flying machines will be possible.
Safety concerns are not major problems, except during landing. Planes must not collide under any circumstances, but could glide or use parachutes if engine failure occurs, or the control systems break. If the plane is out of control, other, still functional, planes will make sure to not collide with the damaged plane. The plane could be equipped with floating bodies for use in emergency landings on water, if this were combined with a water proof passenger cabin, the low weight of the plane, would allow the pilot to survive in it even in rough seas. With these and other safety mechanisms, security while flying is sure to be very good. In addition, the large number of safety mechanisms would positively influence the public’s perception of the plane.
Such aircrafts would cause many changes in human society. People could live in remote, even normally inaccessible places, distributing humanity more evenly. The traffic situation would improve, and fewer roads would have to be built. People could work farther away from where they live, even in other countries (at least in Europe). My dad actually used to do this when I was about 3, he flew to work (which took about 1-2 hours) and then back the next day, so this would clearly be feasible with automatic aircraft.
As with computer-controlled cars, “drivers” would have more free time, because they would not have to control the vehicle.
Criminal use of these planes brings up issues, which can all be solved, and are not worse than with cars. Fleeing from the policy is possible, and eventually any aircraft has to land. At that time criminals can be caught. Crossing borders without landing for inspections could be outlawed; people who break the law could be tracked and intercepted at landing. Planes could of course also have an override that would give control to governments, but these issues could be very controversial, and would not be absolutely necessary.