Abstract:

The study proposes an IoT-enabled smart bridge health monitoring system that uses sensor networks, vision systems, machine learning, and digital twin technology to monitor and schedule repairs. IoT sensors, such as accelerometers, strain gauges, and temperature sensors, monitor vital elements including vibration (0.0-0.9 g), strain (up to 780 µε), and temperature (15-48°C). High-definition cameras can detect surface fractures as small as 2.2 mm. Sending data to a cloud-based processing system using LoRaWAN and 5G is reliable. Analyse data with powerful algorithms. The study employs CNNs to locate picture cracks with 93% accuracy and LSTM models to predict stress over time. These findings can help engineers develop a digital twin of the bridge. This allows them to model real-time structure behaviour, schedule maintenance, and minimise costs. Results show that maintenance costs drop 38% per year and that the F1 score for damage classification is 0.88. This integrated system improves structural reliability, provides early warning systems, and manages new bridge infrastructure flexibly and cost-effectively.