Every vehicle owner driving around town must have at some time experienced the frustration of waiting on a red traffic signal. It may be late at night and you can’t see any other vehicles on the road but still the traffic light controller must cycle through its predetermined sequence, holding you up while you drum the steering wheel with your fingers, waiting for a green. All that could be set to change now that a patent filed last year by IBM has been granted. It describes an ‘intelligent traffic light controller’ using surveillance cameras to monitor traffic flow and a control system using computers employing AI models and algorithms allowing optimum utilization of the signals to reduce waiting times and improve air quality by encouraging more freely moving traffic streams.

Image: USA Patent 9,965,951
 
The patent granted to IBM describes a system using cameras to gather information of the traffic flow, providing vastly more data than the basic vehicle detection sensors buried in the road surface currently in use . The cameras are able to detect pedestrians and vehicles, taking into account weather conditions, prioritizing emergency vehicles on call and detecting vehicles blocking the intersection.
The patent states "The present invention relates generally to the field of traffic control, and more particularly to controlling a traffic signal through cognitive computing that incorporates real time data at an intersection."
A computer system converts the video data streams from the surveillance cameras to calculate the current traffic flow and control the traffic light phase timing.

Admittedly any computer controlled traffic signals could never match a white-gloved traffic control cop at a busy intersection for entertainment value but if AI controlled lights get me home faster while reducing air pollution and fuel consumption then it’s got to be a good thing.