Adaptive Biased Random Walk Algorithm for Ocean Chemical Feature Tracking using an Autonomous Underwater Vehicle

In recent years, technological advancements have led to the development of underwater assets capable of long-term deployment, high maneuverability and hosting a multitude of onboard sensors. Modern Autonomous Underwater Vehicles (AUVs), equipped with chemical tracking capabilities may locate and survey places of interest based on their chemical signature. Such an ability would help in underwater resources and pollution tracking, locating and identification of their sources. This promises to be of great interest to both the scientific and the industrial community. With similar physical and computational constraints, we argue how an AUV in an underwater environment may be compared to a bacterium finding food in its biocenosis. Nature found the novel method of chemotaxis for the survival of the organism. By mimicking this behavior of single-celled organisms, we present a chemotaxis-inspired Adaptive Biased Random Walk (ABRW) algorithm to formulate a navigation scheme for an AUV. Several improvements are proposed to a simple Biased Random Walk (BRW) Algorithm, which includes an adaptive speed-control of the AUV, Quality Factor and data biasing protocols. These features operate on the data collected in-situ by the AUV and aid in the Iocalization of the source. The results are discussed and effectiveness compared with a simple biased Random walk algorithm. Future improvements are discussed.