Events

Learning and Perception for Safe Human-Robot Collaboration in Construction

Collaborative robots are increasingly being deployed on construction sites to assist human workers with repetitive tasks, improving safety, sustainability, and productivity. Because construction environments are highly unstructured, it is often impractical to pre-program robots for specific construction activities. This research therefore focuses on enabling robots to learn construction tasks alongside human workers while maintaining safe collaboration.

The presentation first introduces robot imitation learning and Vision-Language-Action (VLA) methods that enable robots to learn construction tasks from human demonstration data. It then presents 3D single-object tracking techniques designed to monitor surrounding moving objects using sparse LiDAR point clouds. By advancing collaborative learning approaches and 3D perception capabilities, this research aims to develop robots that function as apprentices working safely alongside human workers, helping move construction sites closer to achieving greater sustainability.

Ci-Jyun Liang

Dr. Liang is an Assistant Professor in the Department of Civil Engineering at Stony Brook University. His research interests include human-robot collaboration, computer vision, reinforcement learning, building information modeling (BIM), and digital twins to promote sustainability and safety in future robotized workplaces such as construction sites and manufacturing factories. Prior to joining SBU, he was a Robotics Fellow in the Division of Safety Research at the National Institute for Occupational Safety and Health (NIOSH).