Content area
Public transport facilitates large number people navigate from one place to another; if used efficiently will ease the traffic in crowded cities however because of the fixed schedules, delayed arrival and crowded buses triggers the citizen to travel in private vehicles. This problem can be resolved by efficient and smart public transport scheduling. Existing systems lack real-time data, semantic context, and timing awareness therefore an active scheduling strategy based on sensor data, Artificial intelligence (AI)-based passenger prediction, and time reasoning is required to boost the quality of the services, lower costs, and adapt to evolving city environments. Therefore, this research proposes SPOT-Route (Semantic and Passenger-aware Ontology-driven Temporal Routing), a smart scheduling framework that integrates AI-based passenger detection, semantic reasoning, and behavioral modeling using SHACL and SPARQL. The Public Urban Transport Scheduling System (PUTSS) algorithm is enhanced with two components: the Statistical Data Component (SDC) and the Real-Time Computer Vision Component (RTCVC), which uses YOLOv8 to detect passenger density and anomalies onboard. Sensor data is semantically annotated using SOSAc ontologies and processed through an Answer Set Programming (ASP)-based reasoner. Temporal behavior is modeled using SHACL shapes and SPARQL rules, enabling dynamic decision-making. The system decides whether to skip, maintain, or add bus runs based on congestion and occupancy metrics and the performance of SPOT-Route framework is validated using simulated and real-world data, which resulted in shows a global accuracy rate of 93.2%.