The Future of Traffic
https://www.sciencemag.org/news/2018/11/watch-just-few-self-driving-cars-stop-traffic-jams
The article argues for the validity of self-driving cars. Particularly, the source references a study that used the simulation of roadway events to gauge the benefits of self-driving cars. The researchers found, when simulating traffic conditions with a combination of self-driving and manually driven cars, traffic jams were reduced significantly and the average car speed was doubled. This was most due to the shared awareness that the self-driving cars had for one another. Events that would have led to significant traffic were avoided due to the self-driving being able to slow down in unison.
This topic really stems back to the idea of network traffic and the Nash Equilibrium associated with it. In the current traffic system, everyone makes choices independently of one another to better optimize their time. In this, we have the possibility of Braess’ Paradox in the event of an added path from any one location to the next. This would lead to an overall increase in time from any point A to any point B. This however would not occur if other cars considered more closely the actions of other cars. Self-driving cars would have the advantage of being more actively aware of the paths that other self-driving cars are taking making situations normally resulting in the Braess’ Paradox be less likely.