The progressive deterioration of critical civil infrastructure including bridges, dams, tunnels, and flyovers poses an escalating threat to public safety, particularly in regions with rapidly ageing structural inventory. Conventional periodic inspection methodologies are inadequate to detect incipient structural anomalies before they manifest as catastrophic failures. This paper presents Trestward, a cost effective, IoT based prototype system for continuous, real time Structural Health Monitoring (SHM) of critical infrastructure assets. The prototype integrates an ESP32 microcontroller, Arduino Uno, MPU6050 six axis IMU, SW 420 vibration sensor, and DHT11 environmental sensor. Sensor data is transmitted over IEEE 802.11 b/g/n Wi Fi via the MQTT protocol to a cloud hosted dashboard, where threshold based anomaly detection classifies structural status into three actionable levels: Safe, Warning, and Critical. The prototype was validated on two hand fabricated scaled physical models representing a highway bridge span and a gravity dam. Preliminary testing indicates that the system achieves real time communication and reliable threshold based structural classification under laboratory conditions. As a direction for future work, a supervised Random Forest machine learning classifier is proposed as the next stage AI layer, to be trained on a labelled dataset collected from the operational prototype. The present paper describes the complete hardware design, sensor fusion strategy, communication stack, Structural Health Index (SHI) formulation, and the proposed AI prediction methodology. Trestward is designed for deployment at a per node hardware cost of approximately USD 35, making it suitable for developing nation infrastructure monitoring contexts.
Abhinav Yadav (Sun,) studied this question.