Collaborative Research: CIRC: Planning-C: An Infrastructure for General-Purpose Learning-Enabled Humanoids

Humanoids are robots that mimic human form and function. Such robots can manuever in human-centered environments and handle human tools. This is important for dull, dirty, and dangerous tasks that are unappealing or risky for people, such as encountered in disaster response.

The 2012 DARPA Robotics Challenge demonstrated humanoids mimicking first responder tasks like navigating rough terrain, climbing ladders, clearing debris, breaching walls, turning valves and driving vehicles. A decade later, these outcomes have translated into social benefits. Beyond disasters, people often suffer from tedious and strenuous work on assembly lines.

The car industry is investing in humanoid robots to offset worker occupational injuries. Market forecasters thus see humanoids as a multi-billion dollar industry by 2034. However, current humanoids are still expensive, fragile, and move slowly.

This demands more academic research to advance the state-of-the-art. This planning project assembles the research community to identify what is needed for the next generation of humanoids, ones that are more affordable, rugged, and moves with motions and speeds akin to people.

The outline of this project's activities involves capturing and disseminating the needs of the research community. Three task forces will capture inputs from a diverse research community on (1) electro-mechanical design; (2) software architecture and control systems; and (3) mixed-reality and data-driven learning. These task forces will respectively hold hybrid workshops in universities in Lafayette (Purdue), Boston (Northeastern) and Philadelphia (Drexel).

These workshops bring a diverse community in robotics, computer vision, machine learning, human-robot interaction, VR/AR digital twins, natural language understanding, brain-machine interfaces, advanced cloud and edge computing, high bandwidth communications, algorithmic and communication foundations for advanced operating systems, intuitive programming languages, and trustworthy computing. This process serves to identify both the hardware and software infrastructure the community needs to yield an affordable, durable, and customizable humanoid.

Finally, the task forces will share community inputs at the flagship IEEE International Conference on Robotics and Automation (ICRA) in Atlanta 2025. The net effect will be a comprehensive list of technical design requirements. This will then be leveraged to propose a NEW or Enhance/Sustain (ENS) Medium or Grand infrastructure grant within the next 2-years.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.