Using the Moon as a great laboratory to reach Mars is one of NASA’s grand plans and also a revealing source of problems and challenges to be solved for Mars exploration. One of the most important is deep space communicationsin which Spain plays a vital role where great distances and “periods power outages” which will cut the connection for weeks.
This presents a major challenge for the crews who will one day embark on a months-long journey to the red planet. It’s about inability to resort to NASA assistance on landthe famous control center in Houston (Texas) from where every aspect on board the spacecraft is monitored.
After holding various positions in NASA’s Small Satellite Branch, Andrés Martínez (Program Executive, Earth Independent Operations (EIO) for Mars) is now part of a select group engineers and scientists focused on the Moon to Mars program (From the Moon to Mars, in Spanish). It is one of the latest bets by the US space agency, whose goals include creating systems for astronauts to conduct operations independently of ground control.
Andrés Martínez, Program Director, Earth Independent Operations (EIO) for Mars.
“They asked me to take responsibility for the preparation of all the technology we were developing,” Martínez explained to EL ESPAÑOL – Omicrono on the occasion of the International Small Satellite and Services Forum (SSSIF) in Malaga. “It’s a huge challenge because he’s trying to create systems that support the crew to not have a bad day and they could lose their lives.”
NASA’s ‘Hall 9000’
Comparison with computer HAL 9000 movie 2001: A Space Odyssey it is inevitable. Andrés Martínez himself makes the same reference. “When I got the position, the first thing I did was watch the film,” he admits. “I told the family that I’m now in charge of developing something similar, but this time it’s going to be friendly,” he jokes about HAL 9000’s “rebellious” approach to the team.
Need create an AI for on-board decision-making It is directly conditioned by astrophysics itself, which determines the rules for interplanetary travel. “There is a period of two weeks when the Sun aligns with Earth and Mars, during which time all communication is blocked,” he points out. What is known in space travel jargon as a blackout.
“At the end of last year, we already published the concept architecture of a mission with 4 crew members heading to Mars.” The NASA studied everything necessary step by step and the travel time necessary to get there was determined to be approximately 7 months, with a stay of 79 days before beginning the journey back to Earth.
A recreation of the HAL 9000 control panel
This information is crucial for solving AI problems independently of any connection to the control center. “How many sensors will we need in the various elements of the spacecraft, such as power systems, propulsion, habitability…”. Everything you need maintain the “heartbeat of the ship”.
The goal is to be a computer able to detect and intervene in any case which occurs on the ship. The communication takes about 20 minutes to reach Mars, to which the same number must be added when a reply is sent. This leaves a total delay of 40 minutes between issuing a message and receiving a reply. Too long to make urgent decisions dependent on Houston’s control alone.
“We need a system that analyzes all the telemetry,” he points out. And also that he is able to fix any problem that may arise during the flight. “You have to take control and figure it out”. They will be based on that machine learningalthough Martínez acknowledges that at the moment the program is in a very early period in which they are analyzing the concept before they start training the algorithms.
Training
Several teams of NASA scientists and engineers have already done some research in this line of space flight support. “Already A similarly developed technology is used on the International Space Station”, although aimed almost exclusively at detecting possible problems rather than solving them.
“The problem is that they’re using technologies that have been tested but haven’t They have several decades behind them“Explains Martínez. Some of them even date back to the 80s and 90s. “Considering that we will get to Mars in about 15 or 20 years, we have to start testing new technologies with potential.”
This is the only way they can continue to be valid and mature once the first humans arrive on the red planet. In the shorter term, this type of artificial intelligence will have the ability to make decisions They will test flights to the moon as part of the Artemis program. “The Orion ship carries a type of autonomous system that will support the crew in a major emergency,” says Martínez.
A starship, one of the ships intended to reach Mars, in a test run.
Omicron
Another crucial point of NASA’s artificial intelligence is whether it will have some type of human control, what is called in English man in a loop or a man in a noose. It goes o.a method of verification and approval or rejection the person of the action that the AI suggests.
“These are issues that we still have to review,” comments Andrés Martínez. “Because, How to make a difficult decision?“There are still many decisions to be made and several gaps remain. “But at the end of the day, we have to save the lives of the crew.
You might also like…
Follow topics that interest you
