Geometry-based Human-Centric Robotics

I am interested in the human centric aspects of robotics: how people specify what a robot should do, interpret what a robot is inspecting, verify that a robot has a valid plan, etc. I am interested in geometric approaches to these problems, where we use our knowledge of geometry and optimization to devise solutions. This work is sometimes application-driven, such as considering using robots for inspection, and sometimes problem-driven, such as trying to develop efficient coverage planners.

However, my preferred approach (geometry and optimization) is rapidly being replaced (in research, at least) by learning approaches. And working on meaningful practical, valuable problems requires working at a scale that is challenging for me to sustain. Therefore, I am moving away from robotics projects where I don’t have collaborators.

I remain interested in things like robot inspection – particularly the human elements in automating inspection – but don’t feel like I am in a place to pursue it without partners.