The AIR group has been founded in 2023. Our vision is to enable robot learning for manipulation and handling tasks in logistics. This involves two key research directions: developing continuous autonomous learning capabilities for robots and enhancing their ability to learn from humans.

We are dedicated to developing versatile and flexible robotics systems that can adapt to different tasks and processes without the need for reprogramming by human experts. This flexibility is crucial in dynamic manufacturing environments, where there may be changes in tasks, production requirements, product designs, materials, and suppliers. 

Our research interests include:

• Online learning and adaptation for physical robotic grasping and manipulation
• Sim-to-real domain adaptation
• Lifelong learning, uncertainty-driven learning
• Vision-based deep (reinforcement) learning
• Interactive human-robot learning, Imitation learning