Navigation in a Search Pointer Robot for Victim Detection in Volcanic Eruption Using Hybrid GPS-Inertial Navigation System

Indonesia has more than 127 active volcanoes, creating hazardous environments for Search and Rescue (SAR) operations where poor visibility, unstable terrain, and degraded GPS signals increase risks for human responders. This study presents the design and validation of a low-cost hybrid GPS–INS navigation system with tilt compensation for mobile robots operating in volcanic disaster zones. The system integrates an MPU6050 IMU, QMC5883L magnetometer, and u-blox NEO-M8N GPS module on an Arduino Mega 2560, employing a complementary filter and PID-based heading control to maintain stable and reliable navigation. Field experiments were conducted in simulated volcanic terrain consisting of rocky surfaces, slopes up to 30°, and ash-covered ground under both clear-sky and partially degraded GPS conditions. Repeated trials showed that the proposed system reduced positioning error compared to standalone GPS, which frequently deviated by more than 10 m, achieving improved accuracy of approximately ±4 m. Heading stability was maintained within ±3.5°, while tilt compensation reduced magnetometer distortion from more than 20° to less than 5° on inclined surfaces. These results demonstrate that the proposed approach provides a practical, real-time, and computationally efficient navigation solution for autonomous point-to-point movement and return-to-deploy functionality. Overall, the system offers a reliable and affordable alternative for enhancing robotic navigation in volcanic SAR environments.