Development of an Indoor Position Monitoring System for Rescue Personnel during Emergencies

Indoor positioning systems (IPS) encounter considerable obstacles, particularly the absence of Global Positioning System signals and the complexity of indoor environments, hindering their reliability during emergency scenarios. Infrastructure-based IPS relying on wireless sensor networks and predefined conditions may become inoperative due to fire or explosions. Meanwhile, infrastructure-less IPS using inertial measurement units (IMU) encounter issues such as sensor drift and cumulative errors. This study proposes an integrated solution that combines IMU and laser range scanners to address these limitations. Specifically, a 6-axis IMU comprising an accelerometer and a gyroscope is employed to recognize actions and movement directions, while LiDAR, in conjunction with a step counter, is implemented to construct a step-length estimation model. By fusing these data sources, the IPS is developed to continuously update the position of rescue personnel within a 2D space. Experimental results indicate that the step-length estimation model achieves a mean absolute error of 5 cm and a standard deviation of 4.36 cm per step. Additionally, the action and movement direction classification model attains a perfect accuracy of 100 %. Notably, the overall positioning error of the system remains under 1 m across all trials, demonstrating the efficacy and potential applicability of the proposed method in indoor emergency rescue scenarios.