Soluções Inteligentes para Semáforos

Abstract

The literature review presents intelligent traffic light (ITS) solutions aimed at mitigating urban congestion. A systematic search in WoS, covering the period 2006–2024 and using terms such as “Smart Traffic Light,” “Intelligent Traffic Light System,” “Heuristic,” and “AI,” retrieved 5.618 articles, analyzed with Excel spreadsheets using Methodi Ordinatio, pybibx, and VOSviewer. The bibliometric analysis revealed clusters focused on deep learning, IoT and edge computing, security and blockchain, flow optimization, and fuzzy logic. Combining Excel and Python refined the selection to 351 articles, enabling the construction of publication and citation graphs over time. A significant growth in publications was observed from 2016 onward, although recent works show lower citation averages. Among the 22 core studies, notable contributions included fog/edge frameworks, quantized ML models, hybrid metaheuristics, multi-agent reinforcement learning, and blockchain for adaptive traffic light control. Key challenges remain in legacy system integration, computational overhead, and the need for continuous calibration.