Toward Human-Level
Robotic Teammates R&D State of the Art and Roadmap to the Future

The emergence of agentic AI and the growing influence of neurosymbolic AI make it incumbent on the robotics community to assess their impact on integrating intentionality and reasoning with embodiment and perception.

While advances in machine learning-based, data-driven approaches have expanded the capabilities of robots in perception and control, cognitive modeling relying on symbolic knowledge, commonsense reasoning and ontologically-grounded perception interpretation is widely considered as the necessary complement to the empirical approaches for developing robust, intentional, reflective, trustworthy, explainable, and socially aware robots.

A core R&D question is how best to scale up from task-specific skills to integrated cognitive-robotic agency. The workshop will address neurosymbolic integration, long-horizon and hierarchical planning, human-robotic teaming and theory of mind, perception and action grounding, varieties of learning, trust, explainability, ethics, resource sharing, and other relevant topics.