Search strategies in a turbulent flow using a POMDP framework

When searching for a distant food source, an insect (such as a moth or mosquito) generally cannot rely on chemotactic strategies which climb the concentration gradient of an emitted cue (such as heat, carbon dioxide, or an odor). On the macroscopic scales of interest, turbulence mixes the cue into patches of relatively high concentration over a background of very low concentration, so that the insect will only detect the cue intermittently. In the face of such limited information, locating the source becomes a nontrivial problem. In this work, we cast this search problem in the language of a partially observable Markov decision process and compute strategies that are near-optimal with respect to the arrival time. The trajectories and arrival time pdfs associated with these near-optimal strategies are compared with those associated with a number of heuristic strategies.