project image or logo
Caption/alternate text: Dust Devils on Mars captured by WATCH/SPOTTER Software.
Image credit: JPL/Caltech/NASA


Sometimes scientists would like to capture infrequent events such as Dust Devils on Mars. Or they may be interested in tracking other atmospheric phenomena such as clouds. In these cases onboard analysis can be used to reduce the amount of data that needs to be downlinked, reducing Deep Space Network demands and allowing for additional science.


The WATCH-SPOTTER software was developed for two science applications for the Mars Exploration Rovers (MER) Mission.

First, change detection software was used to detect moving objects such as dust devils in 20 frame movies acquired by the rover. This image processing software could operate in BATCH mode in which all of the images are acquired then analyzed for moving objects or FEED mode in which objects are detected on-the-fly as images are acquired. The portion of the images that does not contain moving objects can then be greyed out resulting in imagery that compresses extremely well, dramatically reducing downlink.

Second, image processing software was developed to detect clouds in sky imagery. The dynamic range of such imagery can then be reduced, reducing image size but retaining the science content of the clouds.

The eventual goal would be onboard summarization software that would downlink dust devil and cloud statistics rather than even reduced imagery.


WATCH-SPOTTER achieved a 75% data volume or 4x data reduction on observed dust devil sequences. Assuming no false alarms (an unrealistic assumption), screening for dust devils also enables a 100% data reduction when the image sequence does not contain dust devils.

The cloud analysis software achieved a well over 90% accuracy rate.


The MER WATCH-SPOTTER software was operationally qualified for the MER mission in 2007.



Jet Propulsion Laboratory, California Institute of Technology
 Andres Castano
 Alex Fukunaga
 Jeffrey Biesiadecki
 Rebecca Castano
 Steve Chien

Arizona State University
 Lynn Neakrase
 Patrick Whelley
 Ronald Greeley

Arizona State University
 Mark Lemmon


This work was supported by NASA as follows:
- algorithm prototyping and testing by the Software Intelligent Systems and Modeling (SISM) program under the OASIS task,
- algorithm refinement and flight code integration by the Mars Technology program (MTP) under the MER On-board Science task and,
- further algorithm refinement and MER testing and operation by the New Millennium Program (NMP) under the MER Infusion Studies task.