Ju-Bong Kim
Papers
3
Total Citations
51
H-Index
3
About
Ju-Bong Kim is a leading researcher at the intersection of reinforcement learning (RL), federated learning, and edge computing, with a focus on real-world autonomous control systems. His work addresses a critical challenge: bridging the "reality gap" that prevents simulated RL policies from being directly applied to physical devices. Kim’s major contributions center on developing **federated reinforcement learning (FRL) acceleration methods** that enable multiple agents—such as robotic arms or rotary inverted pendulums—to collaboratively learn optimal control policies across distributed edge environments without sharing raw data. His 2021 paper on FRL acceleration for precise multi-device control (25 citations) and his 2020 work on controlling multiple inverted pendulums (22 citations) are among his most influential, demonstrating practical solutions for scalable, privacy-preserving automation. He has also extended these principles to SDN-based IoT environments, advancing autonomous control in smart factories and robotics. By enabling efficient, decentralized learning across heterogeneous devices, Kim’s research paves the way for robust, real-time control systems that operate reliably outside the lab—a vital step toward deploying RL in safety-critical, real-world applications.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3