Search NASA⌕ Search
Home/Archive/NASA NTRS

NASA NTRS · 20180008578

Earth Science Deep Learning: Applications and Lessons Learned

Maskey, Manil·Ramachandran, Rahul·Miller, J. J.·Zhang, Jia·Gurung, Iksha
View original source ↗Export metadata ↓

Abstract

Deep learning has revolutionized computer vision and natural language processing with various algorithms scaled using high-performance computing. At the NASA Marshall Space Flight Center (MSFC), the Data Science and Informatics Group (DSIG) has been using deep learning for a variety of Earth science applications. This paper provides examples of the applications and also addresses some of the challenges that were encountered.

Explore related subjects

TrainingDeep LearningSupervised LearningNeural Network

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Maskey, Manil, Ramachandran, Rahul, Miller, J. J., Zhang, Jia, Gurung, Iksha. 2018-07-22. Earth Science Deep Learning: Applications and Lessons Learned. https://ntrs.nasa.gov/citations/20180008578

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related reports

NASA NTRS2018

Applying Research-Based Training Principles: Towards Crew-Centered, Mission-Oriented Space Flight Training

This chapter describes a training approach that applies empirically derived principles of training to re-imagining the overall design of NASA’s space flight training program. The chapter is focused specifically on the design of astronaut training for NASA’s future deep space, exploration missions to Mars. We briefly describe NASA’s space flight training practices during the Apollo and Space Shuttle eras as well as NASA’s current practices for training astronauts for their missions to the International Space Station. We provide an overview of NASA’s current concepts for a mission to Mars to scope our training approach. We envision a new space flight training approach which we term “crew-centered, mission oriented” training, inspired by the design approach offered in the context of airline pilot training by Barshi (2015). We apply some of the training principles reviewed by Kole and his colleagues in the companion volume (Kole, Healy, Schneider & Barshi, 2019), as well as by other researchers in training science (e.g., Ericsson, Krampe, & Tesch-Römer, 1993; Healy & Bourne, 2012; Salas, Wilson, Priest and Guthrie, 2006), into real-world, practical guidelines for the particular context of training astronauts for a mission to Mars.processes over very long retention intervals.

Training↗
NASA NTRS2015

Apparatus and Method for Low-Temperature Training of Shape Memory Alloys

An apparatus and method for the low-temperature thermo-mechanical training of shape memory alloys (SMA) has been developed. The experimental SMA materials are being evaluated as prototypes for applicability in novel thermal management systems for future cryogenic applications. Alloys providing two-way actuation at cryogenic temperatures are the chief target. The mechanical training regimen was focused on the controlled movement of rectangular strips, with S-bend configurations, at temperatures as low as 30 K. The custom holding fixture included temperature sensors and a low heat-leak linear actuator with a magnetic coupling. The fixture was mounted to a Gifford-McMahon cryocooler providing up to 25 W of cooling power at 20 K and housed within a custom vacuum chamber. Operations included both training cycles and verification of shape memory movement. The system design and operation are discussed. Results of the training for select prototype alloys are presented.

Training↗