Krishanu Agarwal
Papers
2
Total Citations
6
H-Index
2
About
Krishanu Agarwal is pioneering the intersection of socially-aware navigation and legged robotics, with a focus on enabling bipedal robots to move safely and naturally in human-crowded environments. His major contributions center on developing real-time, model predictive control frameworks that integrate zonotope-based neural networks—a novel geometric representation for uncertainty—to predict pedestrian motion and plan socially acceptable paths for bipedal agents. In his highly cited 2024 work, "Socially Acceptable Bipedal Robot Navigation via Social Zonotope Network Model Predictive Control" (4 citations), Agarwal introduced a bidirectional coupling between prediction and planning, addressing a critical gap in legged robot navigation. His follow-up study, "Real-time Model Predictive Control with Zonotope-Based Neural Networks for Bipedal Social Navigation" (2 citations), further advanced this by proposing cascaded Pedestrian Prediction Networks for real-time human trajectory forecasting. Though early in his career, Agarwal’s work has already garnered attention for its innovative use of zonotopes—a departure from traditional Gaussian-based methods—offering robust, computationally efficient solutions for dynamic, crowded settings. His research promises to unlock new capabilities for bipedal robots in public spaces, from assistive robotics to autonomous delivery, making him a rising figure in socially-aware locomotion.
Research Focus
Key Achievements
Top Papers
- 1
- 2