Panelist: Joyce Chai Title: Language Grounding towards Situated Human-Robot Communication Transforming language-based communication envisioned by earlier work (such as SHRDLU) to situated human-robot communication faces many new challenges. Robots need to understand objects and related actions, and to perform these actions, all in the physical world. Thus, language needs to be grounded to the perception of the robot, and to internal representations that can facilitate action planning and execution. What makes it even more challenging is that, in situated human-robot communication, although humans and robots are co-present in a shared physical world, they have mismatched capabilities in perception and reasoning. Their representations of the shared world and joint tasks are significantly misaligned. This lack of shared representations makes communication and grounding extremely difficult. To address these challenges, our recent work has applied linguistic knowledge and collaborative behaviors for language grounding. This talk will give a brief introduction to this research effort.