← Back to Research

Building intelligent agents that can interact with and navigate the physical world safely and effectively, enabling robots to perceive, reason, and act in complex real-world environments.

Overview

Robotics and autonomous systems research focuses on creating machines that can operate independently in the physical world. This requires integrating perception, planning, control, and learning to create systems that can adapt to dynamic, uncertain environments.

Our work spans from fundamental algorithms for robot motion planning and control to learning-based approaches that enable robots to acquire new skills through experience and interaction with their environment.

Current Research Focus

Robot Learning and Manipulation

We develop algorithms that enable robots to learn manipulation skills through demonstration, trial and error, and simulation. This includes grasping diverse objects, tool use, assembly tasks, and adapting to novel objects and scenarios with minimal retraining.

Navigation and Motion Planning

Autonomous navigation requires robots to plan safe, efficient paths through complex environments. Our research includes SLAM (Simultaneous Localization and Mapping), dynamic obstacle avoidance, multi-robot coordination, and planning under uncertainty.

Sim-to-Real Transfer

Training robots directly in the real world can be expensive and time-consuming. We investigate methods to train policies in simulation that transfer effectively to real robots, including domain randomization, physics-based simulation, and reality gap bridging techniques.

Human-Robot Collaboration

As robots move from isolated factory floors to human environments, safe and effective collaboration becomes essential. We research intention recognition, shared autonomy, safety constraints, and natural interaction paradigms for human-robot teams.

Breakthrough Applications

• Warehouse Automation: Robots that efficiently sort, pick, and pack items in distribution centers
• Autonomous Vehicles: Self-driving cars, delivery robots, and aerial drones for transportation and logistics
• Healthcare Assistance: Surgical robots, rehabilitation devices, and assistive robots for elderly care
• Disaster Response: Robots that can navigate hazardous environments for search and rescue operations
• Agricultural Automation: Autonomous systems for planting, monitoring, and harvesting crops

Current Challenges

Key challenges in robotics include achieving robust performance in unstructured environments, handling the sim-to-real gap for learned behaviors, ensuring safety in human-robot interaction scenarios, developing more sample-efficient learning approaches, and creating affordable, reliable hardware platforms.

Impact and Future Directions

Robotics and autonomous systems are transforming industries from manufacturing to healthcare to transportation. As AI capabilities advance, robots are becoming more flexible, capable of handling diverse tasks and adapting to new situations.

Future directions include developing more general-purpose robotic systems that can handle diverse tasks, improving energy efficiency and battery life for mobile robots, creating better tactile and force sensing capabilities, advancing multi-robot coordination and swarm behaviors, and ensuring ethical deployment and addressing societal impacts.

Questions about our robotics research? Get in touch