You are here

Humanoid Mobile Robot Manipulation Behavior Development


OUSD (R&E) CRITICAL TECHNOLOGY AREA(S): Trusted AI and Autonomy; Integrated Sensing and Cyber; Human-Machine Interfaces


OBJECTIVE: There is a clear AFSC and DoD need to automate our sustainment processes in order to improve safety, quality, capacity, and readiness. Our near peers are automating at an alarming rate, and we should be developing technologies that keep us as many steps ahead as possible. One limiting factor in the continued scaling/proliferation of automation/robotics is that we have shaped our world to fit the human form factor. And with that, single or dual arm industrial robots stationary or mobile are limited in the number of applications they can support without significant facility or process modifications. These modifications are expensive and often intrusive to process flow, slowing our ability to automate and compete.


DESCRIPTION: Recent developments in the broad capabilities of humanoid general-purpose robotics for item manipulation have positioned these systems to make increasingly significant impacts in sustainment and depot environments by helping to automate human-like activities. These systems can provide improvements in safety, quality, agility, and throughput metrics, allowing overnight or lights-out operation as well as working collaboratively alongside people. For these robotic systems to scale, they must address an ever-growing range of items presented in a vast range of poses and configurations.


One advantage of the humanoid form factor is the ability to adjust the robot’s manipulation workspace, the three-dimensional bubble containing all the points it can touch. In this way, legged mobility critically amplifies bi-manipulation to produce a generalized platform. This platform can then be utilized in an almost unlimited number of ways and applications.

The desired manipulation behaviors are closed-loop to improve their fluidity and robustness, meaning they should be dynamically updated based on real-time multi-modal sensor information (position, torque, vision, etc.) to increase the manipulation speed to near that of what a human normally achieves. The desired manipulation behaviors may require coordination of two arms, including bracing one hand on a surface to enable a long reach or picking up an object using both arms to increase the payload the robot is capable of handling, and utilization of the full body, such as squatting to pick up an object from the ground or leaning back to counterbalance a carried load.


PHASE I: As this is a Direct-to-Phase-II (D2P2) topic, no Phase I awards will be made as a result of this topic. To qualify for this D2P2 topic, the Government expects the applicant(s) to demonstrate feasibility by means of a prior “Phase I-type” effort that does not constitute work undertaken as part of a prior or ongoing SBIR/STTR funding agreement. To demonstrate feasibility, applicant(s) must describe its ability to control humanoid robotic hardware systems to perform a baseline mobile manipulation-based box/item pick. The firm should demonstrate physically, in simulation, or in principle the foundational control algorithms or behavior structures needed to develop workspace configurations for high, low, and far reach.


PHASE II: The objectives of this topic are to develop a working prototype to show increased capability of item picking within a variable workspace; and maximize the efficiency of the robotic system by allowing the robot to operate in a real-world depot environment with items arranged in various locations and positions.


PHASE III DUAL USE APPLICATIONS: If the Phase II is successful in developing the needed technology, WR-ALC will purchase additional systems using organization (working capital) funds. The procurement will include the refinement of hardware and software to increase accuracy and reliability and achieve a production-ready state for procurement by the Air Force, other federal agencies, and private industry.



  1. Jorgensen, S. J., Lanighan, M. W., Bertrand, S. S., Watson, A., Altemus, J. S., Askew, R. S., ... & Mehling, J. S. (2019, October). Deploying the nasa valkyrie humanoid for ied response: An initial approach and evaluation summary. In 2019 IEEE-RAS 19th International Conference on Humanoid Robots (Humanoids) (pp. 1-8). IEEE.;
  2. Jorgensen, S. J., Wonsick, M., Paterson, M., Watson, A., Chase, I., & Mehling, J. S. (2022). cockpit interface for locomotion and manipulation control of the NASA valkyrie humanoid in virtual reality (VR) (No. NASA New Technology Report (NTR): MSC-27278-1).;
  3. Sleiman, Farshidian, and Hutter (2023, August).Versatile Multi-Contact Planning and Control for Legged Loco-Manipulation.;
  4. Chi, Feng, Du, Xu, Cousineau, Burchfiel, and Song. ColumbiaUniversity, Toyota Research Institute (TRI), and MIT. (2023, June). Diffusion Policy: Visuomotor Policy Learning via Action Diffusion.,;
  5. Thibault, Chavez, and Mombaur (2022, November). Standardized Benchmark for Humanoid Whole-Body Manipulation.;


KEYWORDS: Humanoid Mobile Manipulation Robotics; Humanoid Robotics Tool Manipulation

US Flag An Official Website of the United States Government