For almost sixty years, robotic missions have been exploring the surface of Mars in search of evidence of life. There will be more robotic missions in the next fifteen years, the first sample return from Mars will arrive here, and crewed missions will be sent there. Mass spectrometry will be used to analyze samples of the Martian sands to look for signs of past life.
NASA is looking for new methods to analyze geological samples, given how much data we can expect from these missions. The Mars Spectrometry: Detect Evidence for Past Life challenge is being launched by NASA in partnership with HeroX and DrivenData. With a prize purse of $30,000, this challenge seeks innovative methods that rely on machine learning to automatically analyze Martian geological samples for potential signs of past life.
Despite sixty years of concerted efforts by multiple space agencies, the search for life on Mars has yielded little more than incomplete results. Modern surveys show that Mars was a much warmer place billions of years ago. This discovery has led to renewed efforts to find evidence of past life on Mars.
During the Noachian Period. Mars had a denser atmosphere and surface conditions were warm enough that liquid water flowed on its surface. Evidence of this can be found in the form of river channels, sedimentary deposits, delta fans, and other features. Scientists hope to determine how long life could have existed by knowing how long these conditions persisted.
Chemical analysis on soil and rock samples can take a long time. False positives can occur when analyses rely on human interpretation. Scientists hope to automate the chemical analysis process by using machine-learning techniques.
NASA is looking for innovative ways to analyze data obtained by the Sample Analysis at Mars instrument. The data is provided by the NASA GSFC and Johnson Space Center. The SAM instrument has been used for many years to gather soil and rock samples from Mars.
This involves heating samples until they emit gases that can be analyzed by a spectrometer. The SAM instrument has a gas chromatograph that separates gases to aid in identifying them, a mass spectrometer that can detect elements necessary for life, and a laser spectronometer that can detect methane, which is produced by living things.
Greg Lipstein is the Principal of DrivenData.
“This is a fascinating research question where machine learning tools can have a real impact on how we can learn more about our place in the universe. It’s a great chance to harness the collective intelligence and passion of the data community to advance the state of open science.”
According to the Challenge page, the best methods should be able to detect families of chemical compounds. Nitrogen, phosphorous, sulfur, oxygen, and carbon are some of the building blocks of life. Competitors will be able to take advantage of the many experimental runs done on analog samples.
Competitors are tasked with developing machine learning methods that will support scientists in analyzing and interpreting data collected by missions. It is hoped that scientists will be able to conduct future mission operations with greater speed and efficiency. The competition will remain open until April 18th, 2022.
A bonus prize of $2,500 will be given to the winning techniques. The winning entries may be used to help analyze data from Mars and possibly even inform future instruments conducting in-situ analysis. The mission includes the Russian Kazachok lander, the European Space Agency's Rosalind Franklin rover, and the NASA's Dragonfly mission to the largest moon in the solar system.
HeroX CEO Kal K. Sahota said it was exciting to think there might be clues of past life on Mars.
The Challenge is open to anyone over the age of 18 and they can compete as individuals or as a team. If federal sanctions don't prohibit participation, the competition is open to anyone from anywhere in the world. The rules can be found at https://mars.drivendata.org.
Further reading: HeroX.
All Rights Reserved © 2024
