Consortium of institutions receives $6 million from NSF for humanoid robotics research
The National Science Foundation (NSF) has awarded a $6 Million grant to a consortium of institutions led by Drexel University for humanoid robotics research. As highlighted in a recent cover story in Popular Science, there is an overwhelming disparity in terms of resources devoted to humanoids research in the U.S. versus research in other countries that have heavily invested in this area of robotics, such as Korea and Japan. The goal of the new five-year project, part of NSFs Major Research Infrastructure-Recovery and Reinvestment (MRI-R2) program, is to rapidly advance U.S. humanoid research by developing a common open platform.
To date all full-sized humanoids have been individual custom-made units, and advances made using one design do not necessarily translate to others, said Youngmoo Kim, a professor of electrical and computer engineering at Drexel, who is the principal investigator on the project. Currently, Drexel is the only institution in the United States that has HUBO, a world-class adult-sized humanoid developed at the Korea Advanced Institute for Science and Technology (KAIST), resulting from a prior NSF Partnership for International Research and Education (PIRE) award. Building upon this robot, researchers will develop a new common platform (HUBO+) that will consist of the worlds first homogenous full-sized humanoid team, and each of the participating schools will have access to a HUBO+ unit to enhance their research and education efforts.
The universities involved in this award might be able to develop the much needed standards for the humanoids that, indeed, could have an effect close to the one that followed the Internet once it was standardized, said Rita V. Rodriguez, program director, NSF Directorate for Computer & Information Science & Engineering, Division of Computer and Network Systems.
Co-investigators leading the effort are Yury Gogotsi and William Regli of Drexel, Dennis Hong of Virginia Tech and Stefan Schaal of the University of Southern California. Senior investigators from partnering institutions include C.S. George Lee of Purdue University, Yuan Zheng and David Orin of Ohio State University, Russell Tedrake of Massachusetts Institute of Technology, Daniel Lee of the University of Pennsylvania and James Kuffner of Carnegie Mellon University.
These partners represent a critical mass of humanoids research and brainpower, and this effort will, for the first time, enable researchers to work with a common instrument, said Kim. Building upon the unique expertise we have developed at Drexel in assembling and maintaining HUBO, the proposed platform will significantly extend its current capabilities, resulting in six identical units.
As part of the 7-university consortium, Purdue will be receiving a human-size humanoid robot (like a Honda ASIMO), and the Purdue team led by Professor C.S. George Lee (together with Professors Zygmunt Pizlo and Shirley Rietdyk) will focus research on vision-based mobility and human-humanoid interaction.
Current humanoids are rarely autonomous and not ready for unconstrained interaction with humans. Having a consistent platform will facilitate rapid progress in areas needed for autonomy and natural interaction, including mobility, manipulation, vision, speech communication and cognition, and learning.
Humanoids research is inherently interdisciplinary and integrative and interests younger students. The projects outreach partners, including several high-profile museums, will introduce people of all ages to the technologies of robotics, particularly useful in recruiting K-12 students into science, engineering and mathematics.
Other project highlights:
Drexel University will serve as the clearinghouse for information and training (development, construction, and maintenance) for the project. Ultimately, this MRI-R2 project facilitates the following potentially transformative advances in robotics:
- A state-of-the-art, standardized humanoid platform with embedded capabilities for sensing, manipulation, and rapid locomotion, ideal for a broad range of future humanoids research.
- The ability to directly compare and cross-validate algorithms and methodologies and to consistently benchmark results across research teams for the first time.
- A novel energy storage technology for humanoid robotics incorporating supercapacitors for operations requiring high power density, far exceeding the capabilities of traditional battery-only power sources.
- A widely distributed platform that will motivate, recruit, and train a broad range of students spanning multiple disciplines, including artificial intelligence, digital signal processing, mechanics and controls.