(w/m/d) Master Thesis

Detection and mapping of dispersion processes, such as pollutants or gaseous substances in the air, is a very challenging but important problem in a variety of applications, such as monitoring and inspection of industrial facilities or environment protection or pollution monitoring. Typically, stationary sensor networks are used for such tasks. However, stationary sensor networks offer little flexibility in capturing highly dynamic nature of the gases being monitored. The research group Swarm Exploration at the Institute of Communications and Navigation investigates the use of robotic swarms to explore such processes.

Within the Master Thesis the applicant will investigate the use of model-based AI methods, and specifically, hybrid echo state neural networks that are coupled to physics-driven modeling of material propagation based on partial differential advection-diffusion equation. The objective is to capture the spatial dynamics of propagating gaseous material from a single source in 3D under the influence of wind. Both simulative investigations as well as processing of real data collected in wind tunnel experiment are to be addressed within the thesis.