Aniruddha Kalkar
Papers
2
Total Citations
10
H-Index
2
About
Aniruddha Kalkar is a researcher advancing the frontiers of robot locomotion and evolutionary optimization. His key work focuses on developing scalable algorithms that enable robots to adapt online to damage by pre-training diverse, high-performing neural network controllers in simulation. His most cited paper, “Training Diverse High-Dimensional Controllers by Scaling Covariance Matrix Adaptation MAP-Annealing” (2023, 6 citations), tackles a critical bottleneck: the prohibitive cost and hyperparameter tuning required to generate such diverse controllers. By scaling the Covariance Matrix Adaptation MAP-Annealing algorithm, Kalkar’s approach efficiently discovers a wide range of robust control strategies, dramatically reducing the need for expensive network training. This innovation has direct implications for resilient robotics, allowing machines to recover from physical damage without human intervention. His follow-up work (2022, 4 citations) further refines these methods, demonstrating consistent impact in a rapidly evolving field. Kalkar’s contributions are notable for bridging the gap between theoretical optimization and practical robotic resilience, offering a pathway toward more autonomous and adaptable systems. His research is essential reading for students and engineers interested in evolutionary computation, reinforcement learning, and embodied AI.
Research Focus
Key Achievements
Top Papers
- 1
- 2