Skip to main content
The AIG web pages make heavy use of cascading style sheet features for formatting. You may still browse the text of the site, but for best results, please use a CSS enabled browser. Netscape 6 and Mozilla 5 are good. IE 5 will do.

JPL Header

Navigation Sidebar

Main Content

CL 18-2299

HYperSPectral Infra-Red Imager (HyspIRI) Mission Concept Intelligent Payload Module (IPM)

Background

Future space missions will enable unprecedented monitoring of the Earth's environment and will generate immense volumes of science data. Getting this data to ground communications stations, through science processing, and delivered to end users is a tremendous challenge. As an example, the HyspIRI mission concept (in its current design) would generate a staggering gigabit (109 bits), or 1.5 megapixels, per second. Depending on the area and type of the Earth's surface being overflown, scientists and authorities on the ground can use timely information from this unique remote-sensing platform to track a range of terrestrial phenomena.

AI Technology

Problem

cryosphere image classification on hyperspectral data

Moving the functions of event detection and product generation onboard a spacecraft entails several significant challenges. First, instrument data onboard the spacecraft has not yet been calibrated or corrected for instrument or environmental effects. For example, when an instrument onboard a spacecraft images the surface of the Earth, the signal from the surface is distorted as it passes through the Earth's atmosphere up to the instrument. In many applications this distortion must be corrected before the data can be used. Although in some cases simplified versions of this correction can take place onboard, in other cases correction steps require data not available onboard the spacecraft (such as weather, temperature, or other atmospheric data). However, even without such correction, less precise, rapid-delivery products could be of significant interest.

The second challenge is that onboard computing is typically severely limited. Ground-based computers typically are endowed with gigabytes of RAM and gigaflops of computing capability. Typical onboard computing resources for a current spacecraft consist of 64 Mbytes of RAM and processing power of 200 million instructions per second. For example, the Mars Reconnaissance Orbiter (MRO) is flying a Rad 750 processor clocked at 133 MHz.

The third challenge is that current instruments produce enormous amounts of data. A single image from the HiRise camera on MRO is 16.4 Gbits (uncompressed). The HyspIRI TIR instrument produces 1.2 Mpixels per second at eight bands, and the HyspIRI VSWIR produces 300 Kpixels per second at 220 bands. Keeping up with these data rates requires efficient algorithms, streamlined data flows, and careful systems engineering.

Impact

The application areas of these products include flooding, volcanoes, and cryosphere. These applications and many others represent key applications areas for the proposed HyspIRI Intelligent Payload Module (IPM).

Although delivery of the highest-quality products (using standard downlink and processing methods) might be delivered to scientists a week or two after collection, specific applications can benefit from data and product delivery in near real time. For example, active fire mapping can assist in fighting forest fires—enabling better placement and use of scarce firefighting resources.

Status

The HyspIRI Mission Concept has baselined the use of the Intelligent Payload Module concept to bring down rapid response data.

Description

thermal detection at mount erebus from hyperspectral data

The HyspIRI operational concept uses AI techniques in both the onboard and ground mission segments.

On the ground, the spacecraft's orbit is projected, and automated mission-planning tools determine which onboard-processing mode the spacecraft should use. The orbit determines the type of terrain that the spacecraft would be overflying—land, ice, coast, or ocean, for instance. Each terrain mask implies a set of requested modes and priorities. For example, when a spacecraft overflies polar or mountainous regions, producing snow and ice coverage maps can provide valuable science data. The science team can adjust these priorities on the basis of additional information (such as external knowledge of an active volcano, a flooded area, an active wildfire, or a harmful algal bloom). The mission-planning tool accepts all these requests and priorities, then determines which onboard-processing algorithms will be active by selecting the highest-priority requests that fit within the onboard CPU resources, band-processing limitations, and downlink bandwidth.

In the intelligent onboard processing concept, HyspIRI's onboard processing algorithms would consist of expert-derived decision tree classifiers, machine-learned classifiers such as SVM classifiers and regressions, classification and regression trees (CART), Bayesian maximum-likelihood classifiers, spectral angle mappers, and direct implementations of spectral band indices and science products.

Team

Jet Propulsion Laboratory, California Institute of Technology:
Steve Chien
David Mclaren
Daniel Tran
Ashley Gerard Davies
Joshua Doubleday
Goddard Space Flight Center:
Daniel Mandl

Publications

Contacts

JPL Technical Contact: Dr. Steve Chien
Steve.Chien@jpl.nasa.gov
818.393.5320
Software Licensing: http://download.jpl.nasa.gov

Sponsors

Earth Observing One Mission
  • Daniel Mandl (GSFC), Mission Manager
  • Cheryl Yuhas, NASA HQ POC
HyspIRI Concept Team
  • Woody Turner, NASA HQ POC

Footer