Raghavv Goel

I am a masters of robotics student at Carnegie Mellon University where I work on control theory, computer vision and RL with application to surgical robotics under the guidance of Professor Howie Choset and Professor John Galeotti.

Prior to CMU, I was fortunate to collaborate with Dr. Sayan Basu Roy and Dr. P. B. Sujit related to adaptive control, parametric uncertainty, and distributed systems during my undergraduate studies at IIIT Delhi, where I also recieved my deparment (ECE) gold medals for best academic performance and all-round excellence.

I was also part of RISS 2019 program where I collaborated with Professor Katia Sycara on multi-agent task allocation problem.

Email  /  CV  /  Google Scholar  /  Github

I'm interested in coalescing control theory (stability and optimality gaurantees), reinforcement learning (data-driven method) and perception (sensing).

Scanning unknown surfaces (unknown anatomy) is better with sampling the unknown surface at different points and creating an approximate map of the anatomy and surface curvature guided by an information gain

A novel unified method for estimating unknown time-varying parameters with convergence gaurantees

Reference model (leader agent) takes feedback from follower agents but doesn't share the external input with any agent. We design distributed controller, parameter estimator and external input estimator while removing any dependence on global knowledge about the graph laplacian

A novel multi-agent leader follower framework called CRM-DMRAC is proposed where reference model (leader agent) takes feedback from follower agents leading to faster estimation of parametric uncertainty in dynamics and faster converge to external input

We propose the use of a few human influenced agent to control swarms of agents and instill different behaviour in the swarm

We propose to solve task allocation problem in heterogeneous agents using mix integer optimization with collision avoidance and communication breakage constraints