Adaptive control systems for traffic signaling has been one of the most productive and studied research areas in the field of Intelligent Transportation Systems (ITS), as an effective solution to mitigate congestion and avoid its negative effects, and at the same time improving traffic flow and mobility. Nevertheless, the best methods in terms of traffic performance usually present a high computational cost and an unaffordable real implementation and/or go-to-market strategy. With the advent of the Internet of Things, the possibilities of developing new techniques, systems, applications, and services for ITS able to take advantage of the IoT capabilities (e.g., low-cost) will be huge. In this paper, we contribute to the convergence of these two paradigms by introducing PROA, a PROActive intelligent system for optimizing traffic signaling. Based on two modules, i) a new image processing algorithm acting as a traffic-monitoring tool that provides statistical data on vehicle traffic and ii) a novel adaptive traffic signaling algorithm to adjust traffic lights in accordance with the current network conditions, PROA is able to improve key metrics on traffic performance with a low computational complexity.