Development of a wireless sensor network for early detection of sinkholes using ground vibration and soil moisture sensors

Sinkholes are a substantial hazard to society and infrastructure. Conventional detection techniques can be reactive and expensive. This study develops a low-cost, IoT enabled wireless sensor network for early sinkhole detection using ground vibration and soil moisture sensors, integrating an SW-18010P vibration sensor and a capacitive soil moisture sensor with an ESP32 microcontroller and Blynk platform. Real-time data transmission to the Blynk dashboard enabled remote visualization and alerting. In controlled simulations replicating karst dissolution, the system achieved 100% alert accuracy, triggering warnings for combined thresholds such as 69% moisture and 3709 analog read. Cavity formation was simulated using a dissolvable rock sugar setup, which effectively set off alarms when both thresholds were surpassed. Dual-sensor logic, which required simultaneous vibration and moisture anomalies, helped lower false positives. The system is scalable for field deployment in high-risk areas due to its small size, low power consumption, and cost. In conclusion, this study shows that it is possible to use widely available IoT technologies for real-time geohazard monitoring, giving populations at risk of subsurface instability an early warning system.