AI and Machine Learning for Smallsat Data Processing

Smallsats are generating growing amounts of data, via their on-board payloads and in the course of their operation.  Artificial Intelligence (AI) and Machine Learning (ML) are increasingly applicable to smallsats to extract information and optimize performance.

Earth observation and remote sensing applications provide dozens of examples of effective use of AI/ML, easily generating sufficiently large training data sets.  Smallsats now provide imagery at different wavelengths, spatial resolutions and temporal frequency, and the combination of this imagery with other data streams is leading to new AI/ML uses for change detection and more.

Communications payload data processing and routing benefit from AI/ML as traffic is managed over agile distributed networks.  Satellite services must now deliver cloud and intelligent edge applications, reliably and at scale, anywhere in the world, consistent with modern enterprise expectations. New thresholds of capacity and low-latency performance are a given, but just as critical are a range of agile networking capabilities along the principles of the cloud—dynamic bandwidth allocation to the end-user, one step away from the data center, flexibility to support on-demand, consumption-based business models and better visibility and control over network behavior.

AI and ML also apply to the physical operation of satellite fleets in a more automated way that allocates and accesses resources and identifies and resolves anomalies.

Date: February 5, 2020 Time: 1:30 pm - 2:15 pm Dr. Eric Anderson

And One Technologies
Dr. Brian Barritt
Technical Lead / Manager, Loon

Loon (Part of Alphabet)
Marshall Culpepper
CEO & Co-Founder

Marlu Oswald

Seed Innovations
Jake Shermeyer
Senior Research Scientist

In-Q-Tel CosmiQ Works
Rich Waterman
Vice President of Space Systems Integration

Parsons Corporation