Using a mobile robot for hazardous substances detection in a factory environment

Abstract

Industries that work with toxic materials need extensive security protocols to avoid accidents. Instead of having fixed sensors, the concept of assembling the sensors on a mobile robot that performs the scanning through a defined path is cheaper, configurable and adaptable. This work describes a mobile robot, equipped with several gas sensors and LIDAR, that follows a trajectory based on waypoints, simulating a working Autonomous Guided Vehicle (AGV). At the same time, the robot keeps measuring for toxic gases. In other words, the robot follows the trajectory while the gas concentration is under a defined value. Otherwise, it starts the autonomous leakage search based on a search algorithm that allows to find the leakage position avoiding obstacles in real time. The proposed methodology is verified in simulation based on a model of the real robot. Therefore, three path plannings were developed and their performance compared. A Light Detection And Ranging (LIDAR) device was integrated with the path planning to propose an obstacle avoidance system with a dilation technique to enlarge the obstacles, thus, considering the robot’s dimensions. Moreover, if needed, the robot can be remotely operated with visual feedback. In addition, a controller was made for the robot. Gas sensors were embedded in the robot with Finite Impulse Response (FIR) filter to process the data. A low cost AGV was developed to compete in Festival Nacional de Robótica (Portuguese Robotics Open) 2019 - Gondomar, describing the robot’s control and software solution to the competition.