NASA STTR 2007 Solicitation


PROPOSAL NUMBER: 07-1 T5.02-9886
RESEARCH SUBTOPIC TITLE: Algorithms for Autonomous Robotic Materials Handling
PROPOSAL TITLE: Control Suite and Teleoperator Interface for Whole-Body Mobile Manipulators

NAME: digitROBOTICS LLC NAME: The University of Massachusetts, Amherst
STREET: 19 Crestview Dr. STREET: c/o OGCA, Research Admin. Bldg, 70 Butterfield Terrace
CITY: South Deerfield CITY: Amherst
STATE/ZIP: MA  01373 - 1202 STATE/ZIP: MA  01003 - 9242
PHONE: (413) 374-3048 PHONE: (413) 545-0698

PRINCIPAL INVESTIGATOR/PROJECT MANAGER (Name, E-mail, Mail Address, City/State/Zip, Phone)
Roderic A Grupen

Expected Technology Readiness Level (TRL) upon completion of contract: 3 to 4

TECHNICAL ABSTRACT (Limit 2000 characters, approximately 200 words)
Robots that can move about in terrestrial environments and manipulate large and small objects serve a critical role in NASA's Moon/Mars initiative. Such systems will need to serve as precursors to human missions, collaborate with humans on-site, and carry on when humans have departed. This proposal concerns a control suite and teleoperator interface for whole-body mobile manipulators (WBMMs). WBMMs are a class of redundant, mobile, multi-limbed platforms able to use their entire bodies for manipulation tasks. Such platforms are extremely versatile and can address NASA's need for platforms that perform useful work on the surface of the Moon or Mars. To exploit WBMM, a control framework capable of exploiting redundancy and new man-machine interfaces are required. digitROBOTICS and the Laboratory for Perceptual Robotics at UMass propose to build a WBMM control suite and teleoperator interface to address these challenges. We propose to use control basis framework developed at UMass over the past two decades and used to advance the state-of-the-art in many robotic task domains including grasping and manipulation. The WBMM control suite will simplify controller design and allows control knowledge to be easily ported between robots and operating contexts. The WBMM teleoperator interface will allow for varying levels of autonomy, and can preserve safety constraints using control mechanisms from the WBMM control suite.
The anticipated result of the Phase 1 effort is preliminary software for controlling WBMMs as described above at TRL 4, and a proof of concept teleoperator interface at TRL 3. The anticipated result of a Phase 2 effort would be a commercial-grade version of both the control suite and teleoperator interface, and to have the control suite and teleoperator interface running on one of NASA's WBMMs (ATHLETE or Robonaut, for example). The results of the Phase 2 effort are estimated to be at TRL 5-6.

POTENTIAL NASA COMMERCIAL APPLICATIONS (Limit 1500 characters, approximately 150 words)
The proposed software tools can be used to build control software for whole-body mobile manipulators (WBMMs) for planetary exploration, including such tasks as deploying equipment, sample gathering, habitat placement, site construction, deploying cables/conduit, connecting umbilicals, moving regolith, pad construction, and robotic helpers/followers. These software tools can provide existing NASA robotic platforms such as ATHLETE (Ames Research Center) and Robonaut (Johnson Space Center) with: an easy to use environment for programming manipulation tasks, safe and effective teleoperation in the presence of large time delays, the ability to repurpose hardware resources in response to unexpected contingencies, and the ability to transfer control knowledge between robots and operating contexts. These are crucial features for programming robots in the context of space exploration since they will allow a relatively small number of versatile but complex systems to perform a wide variety of tasks without the need for low-level teleoperation.
The control schema used by the proposed software is portable across different robotic platforms. digitROBOTICS LLC plans to sell an economical WBMM bundled with the proposed software tools. The availability of this package could significantly reduce the cost and time required for an independent research laboratory to start producing control code that could be ported to NASA robotic hardware.

POTENTIAL NON-NASA COMMERCIAL APPLICATIONS (Limit 1500 characters, approximately 150 words)
The proposed software tools can be used to build control software for whole-body mobile manipulators for use in healthcare (as assistive devices for in-home or institutional use), in the military (reconnaissance, surveillance, refueling, carrying ordnance, probing environments, sampling soil), logistics/supply chain (loading/unloading at warehouses), and research on WBMMs. digitROBOTICS LLC plans to develop and sell an economical whole-body mobile manipulator (the uBot-5) bundled with a version of the control suite described in this proposal. We believe this will be a very attractive product for researchers working in the field of mobile manipulation and that the availability of such an economical hardware/software combination could potentially increase the rate of progress in the field of mobile manipulation. The software tools described in this proposal could also be sold as a stand-alone product for use on 3rd party robotic platforms. Additionally, since the control schema created by the proposed software tools are portable across different platforms, this hardware/software product would provide an easy way for independent researchers to develop controllers for extremely expensive and/or commercially unavailable robotic platforms used by NASA, other government agencies, or private companies.

NASA's technology taxonomy has been developed by the SBIR-STTR program to disseminate awareness of proposed and awarded R/R&D in the agency. It is a listing of over 100 technologies, sorted into broad categories, of interest to NASA.

Autonomous Reasoning/Artificial Intelligence
Human-Computer Interfaces
Human-Robotic Interfaces
Software Development Environments

Form Generated on 09-18-07 17:52