Advanced Sensor Integration and AI Architectures for Next-Generation Traffic Navigation

Traffic congestion represents an urban challenge that authorities are trying to solve through various means. Current traffic management systems do not solve these challenges, which is why the research presents a new proposal for a traffic optimization system. The proposed solution integrates small-sized equipment (ESP32 equipped with accelerometers, gyroscopes, and cameras), cloud-based AI services (Azure Content Safety), and a multi-parametric analytical framework for real-time navigation. The system uses the Traffic Optimization Algorithm (TOA) proposed by the authors to calculate the Global Route Quality Indicator (GRQIk). It associates each route with a value based on which the degree of optimality is estimated. GRQIk is calculated based on the distance traveled, traffic delays, estimated travel time, road safety, and the individual’s sensitivity. Real-time data are collected using ESP32, with a pothole detection threshold set at 0.8 rad/s. Through the TomTom API, four alternative routes are identified. The performance evaluation showed that GRQIk differentiates route quality, with scores ranging from 26.40% for optimal routes to 100% for the least favorable ones. In addition, Azure’s Content Safety API achieved 100% accuracy in identifying violent incidents and accidents. The limitations of the research concern the small number of images available to test the Content Safety service. The research establishes new approaches for future developments in the field of smart transportation.