Machine Learning and Instrument Autonomy Group

Principal Investigator

Mark Wronkiewicz
Email: Mark.D.Wronkiewicz@jpl.nasa.gov

For many missions, the ability to collect data far outstrips the ability to transmit it to Earth. This is especially true for planetary science missions with targets beyond Mars. For these scientific communities, science autonomy algorithms are an area of interest because of their ability to summarize and prioritize scientific data onboard the spacecraft; they provide a means to maximize the science return of a mission in the face of severe data transmission constraints.

SYNOPSIS



There are several examples of such autonomy systems at JPL including the Ocean Worlds Life Surveyor (OWLS) project for detecting life on Ocean Worlds, Responsive Onboard Science for Europa Clipper, and Content-based On-board Summarization to Monitor Infrequent Change (COSMIC) for Mars exploration. While future missions will benefit from the development of these algorithms, there is not yet a way to systematically deploy and operate multiple algorithms on one mission. This is a vital need as the volume and complexity of data acquired by spacecraft continues to increase.

SYNOPSIS aims to provide a reusable, easy-to-use framework to facilitate deployment of multiple onboard science instrument autonomy (OSIA) systems. It acts to integrate and prioritize data products even when data from multiple autonomously-powered instruments is present. Specifically, SYNOPSIS

1.) Enables multi-instrument prioritization based on a tunable combination of utility and diversity. Utility prioritization favors data products with high intrinsic value (according to the science team) while diversity prioritization favors data products with “new” or “different” content.

2.)Provides a rules-based mechanism to enforce inter-instrument relationships during data prioritization. For example, this is valuable when contextual and high-resolution instruments sometimes measure the same phenomenon.

3.)Is integrated into the core Flight Software (cFS) and F-Prime flight software packages. This lowers the bar for future mission inclusion.

The SYNOPSIS code base is open source and we aim to provide a standard interface for other research groups to improve or reuse science autonomy algorithms or develop new ones to fit within the same framework. Please reach out if you wish to collaborate!

Please visit our public GitHub repository for more information.