NASA Selects Minority-Serving Institution Proposals to Receive Support
The awardees in the first MSI Space Accelerator have been announced after a competitive application process in which they were encouraged to think like startup businesses.
Three teams from minority-serving academic institutions have been selected as awardees in a first-of-its-kind competition to advance NASA’s goals and meet the agency’s needs in the areas of machine learning, artificial intelligence, and development of autonomous systems. Called MSI Space Accelerator, the competition provides funding, training, and mentorship to develop ideas for systems that can operate without human oversight for future science missions in space and on Earth.
The competition is a partnership between NASA’s Science Mission Directorate, its Earth Science Technology Office, the Minority University Research Education Project (MUREP) within the agency’s Office of STEM Engagement, NASA’s Jet Propulsion Laboratory in Southern California, and Starburst, a global aerospace accelerator company based in Los Angeles.
“The MSI Space Accelerator challenge represents an important new partnership between NASA’s Science Mission Directorate and MUREP. The goal is not only to invest in the best ideas from MSIs, but to diversify our supplier base in the long term,” said NASA Associate Administrator for Technology, Policy, and Strategy Bhavya Lal, addressing participants at a “selection day” hybrid event held virtually and in person at NASA Headquarters in Washington on Thursday, June 2.“This should come as no surprise to any of you: Entrepreneurship is the key to successful innovation. It focuses on solving problems that create jobs and drives economic growth.”
During the event, finalists learned their projects will receive $50,000 in funding. In addition, they will enroll in a 10-week accelerator program, operated by Starburst, that will help them prepare to commercialize their proposals by creating a business plan, refining their pitches, and building relationships within the field. The winning teams will also be matched with mentors at JPL, with whom they will work in person during trainings at the Lab later this summer.
Selected by NASA reviewers from an initial round of applications submitted in March, the winning proposals are from teams based at California State University, Northridge (CSUN); Fayetteville State University (FSU) in North Carolina; and University of Massachusetts Boston. Essentially acting as startup companies, the teams got to pitch their ideas to experts and venture capitalists and receive feedback live at the event.
The winning proposals:
- Led by professors Kyle Dewey and Nhut Ho of the university’s Autonomy Research Center for STEAHM, the CSUN team proposed a new programming language, Proteus, that is designed for writing aerospace software and can “serve as a common language between systems and software engineers,” among other innovations, according to the proposal. The team has already received support from MUREP for this project and has collaborated with experts at JPL on a prototype.
- The FSU team’s proposal, called “Autonomous Systems With On-Demand Inference From Perception Pipelines,” would help groups of multiple robots to work better together. For instance, it would seek to maximize science data from a rock scanned from various viewpoints on Mars. Led by computer science professor Sambit Bhattacharya, the team has already received several previous grants from NASA and JPL.
- The team from Massachusetts, led by engineering professor Tomas Materdey, is developing a system of autonomous drones, recharging stations, and sensor nodes “where every component has knowledge of what is happening elsewhere, so to adapt and adjust to the new situation,” the group’s proposal stated.
Opportunities for follow-on funding are anticipated: NASA’s Advanced Information Systems Technology program is planning an open solicitation – with awards of up to $1 million per year – in the summer of 2023 that will include topics related to trusted autonomy, artificial intelligence, and machine learning.
To learn more about the MSI Space Accelerator, please visit:
Source: JPL
