3D Odor Source Localization using a Micro Aerial Vehicle: System Design and Performance Evaluation

Abstract

Finding chemical compounds in the air has applications when situations such as gas leaks, environmental emergencies and toxic chemical dispersion occur. Enabling robots to undertake this task would provide a powerful tool to prevent dangerous situations and assist humans when emergencies arise. While the dispersion of chemical compounds in the air is intrinsically a three-dimensional (3D) phenomenon, the scientific community tackled primarily two-dimensional (2D) scenarios so far. This is mainly due to the challenges of developing a platform able to successfully provide chemical compounds samples of a 3D space. In this paper, a 3D bioinspired algorithm for odor source localization, previously validated in a controlled physical environment leveraging a robotic manipulator, is adapted for deployment on a micro aerial vehicle equipped with an odor sensor. Given the effect that the propellers have on a gas distribution, the algorithmic adaptation focused on enhancing the sensing strategy of the platform. Additionally, two sensor placement configurations are assessed to determine which one yields best sensing results. A performance evaluation in different environmental scenarios is carried out to test the robustness of the implementation. Two different localization systems are used for the performance evaluation experiments to quantify the impact of localization accuracy on the algorithm's outcome.