Balamurugan Ramachandran

Virginia Tech, Intelligent Machines (Sweden)

Papers

Total Citations

H-Index

About

Balamurugan Ramachandran is a robotics researcher whose work bridges imitation learning and underwater autonomy, with a focus on enabling reliable robot behavior in real-world, sensor-limited environments. His most cited work, "Stable-BC: Controlling Covariate Shift With Stable Behavior Cloning" (2025, 6 citations), addresses a fundamental challenge in imitation learning: the compounding errors that arise when a robot encounters states not seen in expert demonstrations. By introducing a stability-aware formulation, Ramachandran’s method reduces covariate shift, making behavior cloning more robust for deployment. In parallel, his paper "Acoustic Localization of Underwater Robots: A Time of Arrival-Based Particle Filter Approach Using Asynchronous Beacon Pinging" (2023, 4 citations) tackles the critical problem of GPS-denied navigation for underwater vehicles. He develops a particle filter framework that fuses asynchronous acoustic measurements to correct sensor bias, enabling more accurate localization in challenging subsea conditions. Together, these contributions demonstrate Ramachandran’s ability to apply rigorous control and estimation theory to both terrestrial and aquatic robotic systems. His work is particularly notable for its practical orientation—addressing real-world failure modes like distribution shift and sensor drift—and has already begun to influence research in safe imitation learning and underwater robotics.