Description

Faced with the excessive problems generated by the poor management of infrastructure in Brazilian cities, the traffic of vehicles on public roads is one of the greatest. An element that is directly related to this issue is traffic signaling which has limitations that end by harming, even more, the circulation of vehicles in extreme situations, such as heavy rains, traffic accidents, road works, etc. Therefore, observing that current traffic signaling systems can not meet the needs of a road model characterized by a high flow of vehicular traffic, the project stands out because it proposes, in an innovative way, a practical, viable, and accurate proposal to solve such problem.

iTraffic comes up with the main objective of reducing as much as possible traffic jams in the main metropolises, acting specifically on the timing of traffic light cycles, adapting them to the road situation in real-time, dynamically, and intelligently: getting vehicles average speed data from Google Maps API on a traffic light net and detecting the corresponding car flow.

With these numbers, the system makes simulations using a genetic algorithm to find a better timing for traffic light cycles. Then, the system sends these results to each iTraffic device through SIGFOX infrastructure, when the data reaches a device, the microcontroller sends these times to the traffic light using NTCIP-Protocol.

These calculated solutions are stored in our database and are constantly analyzed by Google API, ensuring that the macroscopic traffic flow was decreased and making future interventions faster.

Key-words: genetic algorithm; traffic light timing; internet of things; optimization.

Poster
Final Report