Max Robotics — Global Service Robot Directory + US Certification

About

Jonghyun Choi is a robotics and computer vision researcher whose work bridges the physical and cognitive dimensions of autonomous systems. His research spans two interconnected frontiers: robotic 3D scanning and perception for mechanical engineering applications, and embodied AI agents capable of following natural language instructions in complex environments. His most-cited contribution, "Automatic Pose Generation for Robotic 3-D Scanning of Mechanical Parts" (2020, 41 citations), introduced an automated framework for reverse engineering workflows, significantly reducing the complexity of multi-viewpoint 3D scanning — a practical breakthrough for industrial robotics. More recently, Choi has turned his attention to embodied instruction-following agents, developing systems that can reason compositionally about multi-step domestic tasks. His work on multi-level compositional reasoning (2023) and the ReALFRED benchmark (2024) pushes the boundaries of how robotic agents interpret and execute natural language directives in photo-realistic environments. His 2025 research on multi-modal grounded planning further addresses the annotation bottleneck in training such systems by leveraging large language models efficiently. Together, his contributions reflect a consistent commitment to making robots more perceptive, adaptable, and practically deployable across both industrial and domestic settings.