This research project is funded by the Dirección General de Tráfico (DGT) – Ministerio del Interior (Spain) with reference SPIP2015-01780.
Traffic control systems for optimized signaling is one of the most simple and effective solutions to mitigate congestion, improve traffic flow, improve mobility at signalized intersections, and avoid the negative effects of congestion (e.g. CO2 emissions) while ensuring road safety. The design philosophy of the traffic control systems are undergoing a transition from feedback-based (cause or action / response) systems to those ones that anticipate the cause. This trend has been successfully used in packet switching networks for the provision of Quality of Service (QoS), reacting to congestion before it occurs. Following this trend, the PROA project (Proactive and distributed intelligent system for optimized traffic signaling) aims to design a proactive, distributed, autonomous and intelligent traffic control system capable of adapting/optimizing the signaling intervals of several intersections based on real-time traffic conditions and anticipate a possible congestion situation. Traffic monitoring is done through image processing techniques based on subtraction of the background image. This technique has been commonly used in the automatic analysis of video images, especially in applications where the aim is to detect moving objects in video sequences captured by static cameras. This will require a set of cameras installed at specific intersections (or take advantage of existing ones), so that, connected to a digital system on which the image processing techniques are implemented, the system is capable of providing both the number of vehicles stopped at the intersection and the arrival rate to it. The update interval signaling process will be done through the use of QoS techniques, successfully used in packet switched networks, extrapolated to vehicular traffic and adapted to the characteristics of such networks. To the knowledge of the members of this research team, the translation of these algorithms used in telecommunication networks has been little studied to date but with promising results. Furthermore, in PROA, a group of intersections (formed by one or more control signals – traffic lights -) will form a network under an architecture of distributed system, where decisions are made based on both local and neighboring intersections information. Specialized protocols proposed in the related literature in order to reach consensus in a distributed system (in terms of computer networks) are equally applicable in this new scenario. Also aim of this project is to implement a prototype of the proposed solution for a first experimental study. The team will make an effort to propose a solution enabling the economic viability of the implementation of these devices in the short term.