Abhimanyu Bhowmik

Advancements in robotics demand systems that seamlessly collaborate with humans in dynamic environments. For robots and humans to work well close to each other, robots need to be smart and aware of their surroundings. They should be able to see and understand what people are doing, use what they know, and choose their next steps based on the situation. Previous work has explored isolated aspects of human-robot interaction (HRI), such as safety in motion planning, multimodal interaction with the environment, or human intent recognition. However, a critical gap exists in integrating these aspects into a unified framework. This proposal addresses the challenge of designing a human-robot collaboration system that enables the robot to understand the environment and the human intent from multimodal sensory inputs (e.g., vision, proprioception, audio), while contributing toward shared and individual tasks without jeopardizing safety or efficiency.