I am currently scaling the iLQG belief space variant presented in Stochastic Dynamic Games in Belief Space (arxiv). This involves a generalizable way to select important agents in a multi-agent system. My most recent work has been published as a journal paper to AISy.

I am also interesting in interactive multi-agent planning with low computational power. The goal of this research is to find distributed methods to guarantee better task fulfillment over an environment, where agents can move between static/mobile depots to recharge and report new information if necessary. My most recent work is under review for RA-L.