Martin Asenov (cohort student representative)
Sensors are routinely mounted on robots to acquire sparse measurements of spatio-temporal fields. Locating features within these fields and reconstruction (mapping) of the dense fields can be challenging in many situations, such as while locating the source of a gas leak. In such cases, a model of the underlying complex dynamics can be exploited to discover pathways within the field. We use a fluid simulator, as a model, to infer the location of gas leaks. We perform localization via minimization of the discrepancy between observed measurements and gas concentrations predicted by the simulator. Our method accounts for dynamically varying parameters such as wind and its effects on the distribution of gas. We develop algorithms for offline inference as well as for online path discovery via active sensing. We demonstrate the efficiency, accuracy and versatility of our algorithm using real experiments conducted in outdoor environments. We deploy an unmanned air vehicle (UAV) mounted with a CO2 sensor to automatically seek out a gas cylinder emitting CO2 via a nozzle. We evaluate the accuracy of our algorithm by measuring the error in the inferred location of the nozzle.