Managing oversaturated traffic intersections

An article in today’s Chicago Tribune (Synchronized signals no match for ‘oversaturation’ by Jon Hilkevitch) addresses the issue of traffic lights and capacity. The issues should be nothing new for OR types: synchronizing traffic lights and developing “smarter” traffic lights isn’t very effective when there is just too much traffic.

Having lived in Chicago, I frequently complain about the horribly synchronized traffic lights in Richmond, which are vastly inferior to traffic lights in just about every other city I have driven in. However, the traffic in Chicago is worse than the traffic in Richmond–better traffic lights only exist in cities with terrible traffic. I’d choose less traffic over better traffic lights.

Hilkevitch writes:

The effect [of synchronization] on traffic is nothing short of fabulous, reducing travel times, fuel consumption and pollution — but only when there is available capacity on the road.

Those same interconnected traffic signals are “completely useless” against heavy traffic… The term that traffic engineers use is “oversaturation.” It means there are simply too many vehicles for the available pavement.

Yet the reality of supply and demand isn’t stopping transportation officials at the state, county and municipal levels from spending hundreds of millions on interconnecting traffic signals so they “talk” to each other in an attempt to maximize traffic movement.

This is a lucrative research area: The Transportation Research Board is offering a $600,000 research grant for developing better traffic light synchronization in oversaturated conditions (the deadline is October 16).

This makes me wonder if these approaches do indeed “maximize” capacity. In some situations, very long lights help traffic flow although sometimes drivers have long waits. I rarely see these kind of lights, so maybe drivers want some kind of “equity” (capacity should be reduced so no one has too wait too long). In some areas, lights are so close together that the resulting finite queues present challenges in making smarter traffic lights.

I’ve always wondered about roundabouts, and whether we should use them more frequently in the US. Apparently, roundabouts improve capacity and reduce accidents, but revamping many intersections in a city to create roundabouts would be an expensive, logistical nightmare.