I wonder if the people doing this study wrote into their program the traffic jams caused by the occasional collisions that this type of behavior would tend to create in the real world? When 2 cellular automata "close collide" with each other in a computer simulation, they normally just keep on moving in a new direction. It wouldn't bring them to a halt unless a provision to do so was written into their code. But when 2 vehicles bump each other in real life, that can bring the whole column of traffic to a halt until the collision was cleared. This is one of the pitfalls that can occur when trying to interpret data generated in an idealized world and applying it to the real world. You have to think of and include all of the variables.
|