Estimation of Gas Source Location from Fluctuating Readings of Gas Sensors and Anemometer on Mobile Robot in Outdoor Environment

Abstract

Here we report on a new method for robotic gas source localization. Research efforts are being made to develop mobile robots that can autonomously search for gas sources. The tasks envisaged for such robots include detecting gas leaks in industrial plants and exploring gas emissions in landfill sites. If a mobile robot is provided with gas sensors, it can detect the presence of a target gas drifting in the air. However, to find the location of the gas source is not a trivial task for the robot. Gas molecules released in the air are carried by the airflow and form a plume in the downwind direction. Since the airflow the robot encounters in gas source localization applications is almost always turbulent, the plume randomly meanders and comes to have a complicated patchy structure. Gas source localization in outdoor environments is especially challenging owing to the large fluctuations in the wind direction. In the method proposed in this series of work, a mobile robot equipped with gas sensors and an anemometer is used to measure the gas concentration and airflow direction at various locations in the area under consideration. The location of the gas source is estimated by applying a turbulent diffusion model of a gas plume to the sensor data collected by the mobile robot. Two different turbulent plume models were compared regarding the accuracy in the estimated gas source location. The results are presented to show that the proposed method can provide a map showing the area in which a gas source is likely to be found.