Jose Marti student team works on project with Jet Propulsion Lab

A student team from Jose Marti MAST 6-12 Academy (JMMAST) led by teacher Erica Henderson recently participated in a project to develop an advanced spectrometer processor chip with Dr. Adrian Tang of NASA’s Jet Propulsion Laboratory (JPL) in Pasadena, California.
The development project, which is supported under NASA’s 2018-2021 Astrophysics Research and Analysis Program (APRA), seeks to develop an advanced spectrometer chip for space instruments that can detect trace gasses and chemicals in atmospheres or distant stellar objects. The chip is being developed primarily for use in ground and space-based telescopes to study star-forming regions both in our galaxy and others throughout the visible universe.
Beyond astronomy, the chip is expected to play an important role in planetary exploration of our solar system. The chip provides a means for deep space robotic missions to study the atmospheres of planets and primitive planetary bodies (comets and asteroids) throughout our solar system and detect important compounds like water, oxygen, carbon dioxide and organic chemicals that give us a better understanding of where life might exist and how our solar system formed.
The spectrometer chip itself is an extremely complicated device containing over 200 sub-systems and well over a million transistors. The chip houses complicated digital and analog functions including a very high-speed processor and an ultra-wide band analog to digital converter.
A key challenge that arises from this extreme complexity is how to keep all these different sub-systems within the chip synchronized. Being synchronized ensures that no “traffic jams” occur within the data pathways and that data needed by each sub-system arrives on time, sort of like how Amtrak manages their railroad.
To allow this management, the spectrometer has over 200 different settings that can speed up or slow down various parts of the chip in order to speed up data running late and slow down data that is ahead of schedule. However, there are simply too many things going on at once within the chip for a human being to keep track of and adjust all 200 settings, so an automated approach is needed.
This is where the JMMAST team takes over. Henderson and her student team led by Elizabeth Berenguer and Hadrian Gonzalez Castellanos developed a complex calibration algorithm that autonomously looks at data output by the spectrometer chip and determines if the synchronization is correct, or if settings need to be adjusted.
Detection of synchronization issues from the output data is not a simple problem to solve as the data itself can look vastly different depending on what the spectrometer is looking at, for example Earth’s N2/O2 atmosphere versus the CO2 dominated atmosphere of Venus or Mars. Their algorithm needed to take this into account extracting features from the data that can be attributed to synchronization while not being confused by changes in the underlying data.
As part of the project Berenguer and Henderson visited NASA’s JPL on June 21 to present their results and discuss how it will be integrated into the software that runs on the spectrometer chip. As part of that visit they also enjoyed a tour of the JPL facilities including the spaceflight operations facility, spacecraft assembly facility, in-situ rover laboratory and several of the laboratories where the spectrometer chip is being developed.