Professor Agung Julius's lab has received an NVIDIA Academic Research Grant to support a groundbreaking project titled Concept-based Foundation Models for Vision-Language-Action Planning & Reasoning. The research is based on the innovative work of ECSE Ph.D. student Yunshi Wen.
The project aims to develop robots that understand and perform tasks more like humans. Instead of executing rigid, pre-programmed commands, these robots will act based on high-level "concepts", such as the concept of "putting an object into a drawer" rather than following a precise motion script. By leveraging vast training data and simulated trial-and-error learning, the team seeks to build versatile robots capable of adapting to unfamiliar tasks and environments. Applications range from manufacturing and logistics to assistive robotics.
This prestigious recognition from NVIDIA underscores ECSE department’s expertise in advancing robotics and AI research.