Calibration and Performance Evaluation of Cost-Effective Capacitive Moisture Sensor in Slope Model Experiments.

Area of Science:

• Geotechnical Engineering
• Environmental Monitoring
• Sensor Technology

Background:

• Slope failures, particularly rainfall-induced landslides, pose significant risks.
• Accurate soil moisture monitoring is crucial for landslide early warning systems.
• Existing sensors often require expensive calibration for specific soil types.

Purpose of the Study:

• To evaluate a low-cost capacitive soil moisture sensor (SKU:SEN0193) for landslide risk assessment.
• To develop soil-specific calibration equations for silica sand under artificial rainfall conditions.
• To integrate the sensor with an Internet of Things (IoT) automatic monitoring system.

Main Methods:

• Laboratory-based soil-specific calibration of the capacitive soil moisture sensor.
• Development of calibration equations using polynomial functions.
• Validation of the calibration through slope model experiments with artificial rainfall.
• Integration with an IoT system for automatic data acquisition.

Main Results:

• Polynomial functions provided the best fit for sensor calibration, with R² values from 0.918 to 0.983.
• Root Mean Square Error (RMSE) ranged from 1.171 to 2.488, indicating good model performance.
• Validated sensor readings showed an approximate 12% deviation from actual soil moisture content.

Conclusions:

• The calibrated low-cost capacitive soil moisture sensor demonstrates potential for accurate landslide early warning systems.
• Soil-specific calibration is essential for reliable performance of these sensors.
• The integration with IoT enables cost-effective, widespread deployment for slope stability monitoring.