Hosik Chae
Papers
9
Total Citations
211
H-Index
7
About
Hosik Chae is a roboticist whose research bridges perception, planning, and control for legged and manipulation systems. His most impactful work centers on quadruped locomotion, where he developed the ALPHRED robot—a multi-modal platform for package delivery that has garnered 91 citations—and pioneered a framework for stable, autonomous locomotion over unknown terrain using visual feedback and mixed-integer convex optimization (15 citations). Chae has also made significant contributions to buoyancy-assisted legged robots, notably BALLU2, a lightweight biped designed for safe human interaction, and introduced a residual physics learning method for sim-to-real transfer on these platforms. In manipulation, he proposed an online task-planning framework using mixed-integer programming for dual-arm cooking robots (16 citations), enabling efficient scheduling of complex multi-step tasks. His practical engineering insight is evident in his analysis and noise modeling of the Intel RealSense D435 depth camera (61 citations), a widely used sensor in mobile robotics. Chae’s work is distinguished by its focus on real-world deployment, as shown by his lessons-learned paper on a hotel concierge robot, and by his creation of the Surface Material Dataset for Robotics Applications, which provides friction coefficient data for vision-based terrain segmentation.
Research Focus
Key Achievements
Top Papers
- 1
- 2Analysis and Noise Modeling of the Intel RealSense D435 for Mobile Robots61 citations · 2019
- 3
- 4
- 5
- 6BALLU2: A Safe and Affordable Buoyancy Assisted Biped8 citations · 2021
- 7
- 8
- 9