An approach with ground breaking potential for motion control of completely autonomous vehicles is pursued in this project, which allows to seek, find, and follow a source of a contaminant field using only the measurement of the concentration of the contaminant and not requiring GPS or INS information. This approach builds on recent advances by the PI in the area of real-time optimization via extremum seeking control. The work focuses on the following tasks: (A) extending the initial discovery presented in the proposal from 2D locomotion for mobile robots to 3D locomotion for underwater and aerial vehicles; (B) performing a convergence study for the schemes using averaging and singular perturbation techniques; (C) numerically testing the results in turbulent (rather than merely diffusive) fluid fields; (D) testing of "contaminant cleaning" abilities of the schemes where the contaminant concentration is reduced temporarily along the path traveled by the tracker by releasing an agent that neutralizes the contaminant; (E) extending the results from a single contaminant source and single tracker to multiple contaminant sources and multiple trackers; (F) performing experiments with a mobile robot in an optical field emulating a contaminant field; (G) performing experiments with concentration sensors (for smoke) in an environmental chamber.
