Research

Allen the PR2

My research interests broadly lie at the intersection of robotics, planning and statistics. I am interested in the problem of learning from experience, i.e., building systems that are capable of improving performance by observing their operation. For this, I look for approaches having strong mathematical and theoretical properties.

Currently, I am working on heuristic-based graph search algorithms that exhibit such a behaviour. Prof. P.P. Chakrabarti is supervising my work on a new heuristic-based search algorithm that makes efficient use of islands. Under the guidance of Prof. Bibhas Adhikari, I am also exploring the applications of algebraic topology, particularly those of homology groups in graph search.

In the past, I have spent two wonderful summers in the Search Based Planning Lab at CMU working under the guidance of Prof. Maxim Likhachev on heuristic-based motion planners for robots such as the PR2 and humanoids.

Conference Publications

  1. Learning to Avoid Local Minima in Planning for Static Environments
    Shivam Vats, Venkatraman Narayanan, Maxim Likhachev
    International Conference on Automated Planning and Scheduling (ICAPS), Pittsburgh, 2017