Development of real-time path-based 3D dose rate mapping system

Abstract

According to the ALARA principle, workers in radiation field must use dosimeters at working sites to ensure that annual effective dose does not exceed 50 mSv. However, traditional dosimeters assess radiation levels only after post-processing, which lack a real-time monitoring system. To overcome this limitation, a real-time path-based 3D exposure dose mapping system was developed by integrating a survey-meter and a 3D scanner using the SLAM (Simultaneous Localization and Mapping) algorithm. This advanced system allows simultaneous spatial scanning and dose measurement, enabling free movement in complex indoor/outdoor environments. As workers navigate the area, the system generates a point cloud dataset (PCD) of the environment, recording that coordinates and measured dose rates. This dataset is visually presented in real-time, following the worker's path in 3D space. Additionally, a Deep Neural Network (DNN) model was created to produce a 3D dose rate distribution map. By using the path coordinates as input and corresponding doses as output, the model predicts dose rate throughout the entire PCD. These predictions were used to create a 3D map, with color and brightness adjusted based on dose rates. The system was implemented with LiDAR and a Geiger-Müller detector, then successfully tested in preliminary experiments.