IBM patent describes AI-managed traffic system
The patent – titled “Cognitive traffic signal control” – describes a traffic management system where a computer receives a streaming video for one or more paths of traffic and pedestrians at an intersection. Based on the analysis of the real-time flow of traffic, the computer processor(s) would then determine the best way to manage the traffic signals.
“The computer determines whether a change in a state of a traffic signal for the first intersection should occur based at least in part on the identified traffic and the determined traffic flow with respect to predefined objectives,” the patent reads. “Responsive to determining the change in the state of the traffic signal for the first intersection should occur, the computer calculates a change to a traffic signal timing based on the determined change in the state of the traffic signal. The computer initiates an adaptation to the traffic signal timing based on the determined change to the traffic signal timing.”
A related image shows the cognitive traffic system connected to a variety of systems and devices via a network(s), including “weather sensor,” “camera system,” “vehicle sensor,” and a “traffic signal controller.” The latter is described as “a microprocessor or computer which monitors and alters the operating conditions of a traffic signal.”
The “camera system” is a video surveillance system using one or more video cameras to monitor an intersection where a traffic signal is present. The “weather sensor” is a device or service that monitors the real-time weather conditions at a known location, while the “vehicle sensor” refers to a sensor installed on a vehicle that reports conditions related to the operation of the vehicle.
Related articles:
Traffic sensor market forecast to 2023
Intel-sponsored study on smart cities focuses on citizen benefits
Las Vegas bets on Smart City tech
Wrong-way driver detection system is first in U.S.
Sigfox, PNI Sensor partner on smart parking in the U.S.