Thursday, September 2, 2010

The Wisdom of Crowds in Transportation Networks and the Braess Paradox

Is there a phenomenon of wisdom of crowds in transportation networks?

In the paper, "The Negation of the Braess Paradox as Demand Increases: The Wisdom of Crowds in Transportation Networks," just published in Europhysics Letters, volume 91, (2010), I establish that this is, indeed, the case.

Drivers act selfishly and independently in choosing their optimal routes of travel with no concern as to the effects of their choices on the effects of others or on congestion, which is known, in economics, as a negative externality. Such selfish behavior can lead to the infamous and counterintuitive phenomenon known as the Braess paradox in which the addition of a new road can actually increase the travel time for all! Hence, the removal of a road to traffic as has been done recently in Seoul, Korea and in NYC may actually improve the travel time.

Our research proves that even in networks in which the Braess paradox occurs that, at higher levels of demand, e.g., with more drivers on a transportation network, the new route will no longer be used, and, hence, the Braess paradox will be negated. This novel result suggests that there is a wisdom of crowds phenomenon taking place.

Since other networks, including the Internet and electric power generation and distribution networks, have features similar to that of congested urban transportation networks, this paper has important applications to the design of such networks, as well. Notably, it demonstrates that designers of critical infrastructure must capture the demands and the behavior of the users of the infrastructure in their designs since inappropriate designs may not only result in worse-off situations for all but, ultimately, in a lack of use of parts of the new infrastructure. Fascinatingly, it may also have applications to targeted cancer therapy.

These results are especially timely given the calls for increased investments in infrastructure in the US. Without appropriate planning, however, the so-called investments may actually be counterproductive. At the same time, our study demonstrates that 'if one builds, it does not mean that more will come!'