April 4, 2017, Mountain View, CA – NASA Frontier Development Lab (FDL) announced its 2017 challenges and opened the participant application process, with a deadline of April 19, 2017.
Hosted in Silicon Valley by the SETI Institute, FDL is a graduate-level summer workshop developed in partnership with NASA Ames Research Center. The eight-week program tackles knowledge gaps in space science by pairing machine learning expertise with astronomy and planetary science expertise at the PhD level. Interdisciplinary teams address tightly defined problems and the format encourages rapid iteration and prototyping to create outputs with meaningful application to the space program.
The workshop features partnership with technology leaders in machine learning from the private sector and academia. Partners bring advanced hardware, software and cloud-based resources, and provide coaching to research participants. FDL Partners for 2017 include IBM, The University of Southern California Machine Learning Center, Autodesk, NVIDIA and Miso Technologies. Entering its second year, FDL strives to create research outcomes that support NASA goals, while simultaneously showcasing cutting-edge partner capabilities in deep learning and artificial intelligence.
“We are excited about building on the success of last year’s program, which validated the benefits of applying deep learning capabilities to space science problems, and reinforced the efficacy of public-private partnership.” said Bruce Pittman, Chief Systems Engineer, NASA Ames Space Portal.
The 2017 program expands the size and scope of the 2016 workshop, increasing the number of participants and expanding the research categories to three key topics: Planetary Defense (being smarter than the dinosaurs); Space Weather (predicting the trillion-dollar storm), and Space Resources (living off the land as we build a new frontier).
“From predicting solar flares to accelerating the risk analysis of asteroids in the solar system, cognitive technology delivered on the cloud has the potential to solve some of the most pressing challenges in space science today,” said Mac Devine, IBM Fellow, Vice President and CTO, Watson Cloud Division. “IBM will work with researchers and data scientists in the FDL workshop, to apply AI, machine learning and deep learning techniques on the IBM Cloud to analyze data and rapidly iterate and prototype solutions."
“FDL acknowledges that the brightest talent in AI right now is coming out of academia and the private sector.” said FDL Director, James Parr. “Our vision is to collide these super-smart individuals and organizations with the fascinating problems and rich data resources offered by NASA and create meaningful solutions useful to the space program.”
“The premise and structure of FDL are a perfect fit with our mission,” said Bill Diamond, CEO of the SETI Institute. “It is a privilege to host a research accelerator that leverages our science resources and promotes STEM education programs. Engaging academic and private partnerships further enhances the value of the program. We are delighted to be working with NASA and our partners to host FDL 2.0”
In 2016, 3 teams comprised of data and planetary scientists, and supported by subject matter experts, focused exclusively on planetary defense, covering asteroid shape modeling from radar data, meteorite recovery with machine-vision equipped drones, and asteroid deflection strategies leveraging 1.5 million simulated impact events for training.
A resulting paper, “Application of Machine Learning for Planetary Defense – Three Case Studies,” was selected for presentation at the IAA Planetary Defense Conference in Tokyo and Nvidia’s GPU GTC conference in Washington DC.
To learn more about FDL, the 2017 challenge questions, and to apply, please visit the FDL website at www.frontierdevelopmentlab.org.