Abstract
This study proposes an artificial intelligence-driven smart city system (AISSC) for trend detection, short-term forecasting, and decision support using continuously replayed urban data streams. The framework integrates sliding-window feature generation, adaptive statistical trend detection, forecasting, and reinforcement-learning-based policy recommendations. AISSC was evaluated as a controlled prototype using publicly available benchmark datasets replayed chronologically to emulate streaming operation; it was not deployed in a live city-wide infrastructure. Under the evaluated conditions, AISSC achieved 98.6% directional accuracy, 6.3% MAPE, and 7.8% WMAPE. The mean trend-detection latency was 0.83 s, while the mean end-to-end processing latency was 2.40 s. Dashboard information was refreshed every five minutes, representing the configured update frequency rather than processing latency. Results show the feasibility of AISSC for low-latency urban stream analysis, but more deployment-scale and longitudinal validation is needed. These results demonstrate the capability of the proposed framework to support reliable real-time trend analysis and short-term forecasting for smart city decision-making. This shows that the framework can reliably identify trends and accurately estimate demand for real-time smart city decision-making. The results demonstrate consistent performance improvements compared to representative baseline methods under identical streaming and computational constraints. The proposed AISSC facilitates real-time observation of urban dynamics and enhances short-term forecasting accuracy for informed decision-making in smart city management systems.