오성회 연구실

Develop efficient, safe, and socially friendly machine learning so that autonomous robots can coexist with people in various environments. Includes reinforcement learning with fewer data, safety assurance, and applications in delivery robots and housekeeping robots.

Develop deep reinforcement learning for metaverse applications using digital twin technology. Enable solving real-world problems in smart factories, smart cities, logistics, and robotics through virtual world simulations.

Develop technologies leveraging large language models to enable robots to achieve human-level decision making in diverse situations by solving complex and long-term planning problems.

Develop general-purpose reinforcement learning technology for industrial and service robots in unstructured, complex environments. Utilize human-robot interaction and base models for quick adaptation to new purposes.

Develop cutting-edge technologies for smart logistic platform including order management, linked delivery systems, autonomous delivery vehicles, and real-time fleet management.