A Micro Quadrotor Gas Source Localization Benchmark in a Low-Computational Controller

This work aims to compare various olfactory localization methods in a micro quadrotor equipped with a low-computational controller. Finding the source of an odor is a challenging engineering problem that can be simplified with the use of a high-computational controller. In most cases, a regular quadrotor will suffice to complete an odor localization task. However, a standard-size quadrotor cannot guarantee a successful search when the circumstance arises in a congested area due to space limits. A small quadrotor, such as a micro quadrotor, was extensively used due to the limited space. A small quadrotor appears to be a promising option because it enables rapid mobility and maneuverability in congested areas. On the other hand, a small quadrotor is constrained by considerable constraints, such as payload limitations, and the high-computational controller frequently requires significant dimensions that are difficult to include in a small quadrotor. A low-processing controller is utilized in this situation to assist the odor location search. The source localization technique was developed using the silk moth’s programmed behavior, which includes motion modification, noise reduction with the Savitzky-Golay filter, and estimates with Autoregressive Exogenous. Experimentation confirmed that the micro quadrotor’s odor source localization gave a suitable outcome.