NASA and other agencies already apply artificial intelligence (AI) technologies for greater efficiency and better quality results. The analysis of new craters on the surface of Mars and an AI that scans telescope images to find exoplanets (planets outside the solar system). These are just a few examples of this intersection between space exploration and artificial intelligence, which years ago would sound like science fiction. Below, you can see six situations where NASA used AI to help in its missions.
1. Discovery of new craters on Mars
One of the new applications of artificial intelligence techniques adopted by NASA involves the classification of the images generated by the MRO probe, which has been orbiting Mars since 2006. The tool analyzes thousands of photos to detect surface changes that may indicate the presence of new craters.
The technology uses machine learning and interprets signals that can crater. For this, the AI was trained by NASA with 6,830 different images taken by MRO. In this medium, there are pictures of clean areas of craters and points marked by a meteor impact. Thus, it is possible to create a clear distinction that facilitates recognition, allowing the identification of new points of interest on the Martian surface.
The importance of the tool in the process of studying the neighboring planet lies in relieving the workload of scientists since the AI is more agile in the task.
2. Analysis of Exoplanets
Kepler is the name of a telescope that operated in Earth’s orbit from 2009 to 2018. Its function was to search distant stars in search of potential exoplanets, which are the planets that inhabit solar systems different from ours. As the volume of information generated by Kepler was enormous, the use of artificial intelligence turned out to be a powerful tool in the process of filtering candidates.
The action of AI here is to analyze these data to try to identify the origin of oscillations in the light intensity of distant stars. In this way, technology can differentiate if they are caused by some kind of phenomenon different than the exoplanet traffic. The success of the technique has already yielded the confirmation of 50 new planets in the original data volume of the Kepler mission.
To arrive at this result, the scientists used a collection of images already analyzed by Kepler, containing both real and false-positive planets. Over time, AI has developed the ability to separate one result from another and is now able to perform more efficient analysis in future missions.
3. Projection of new Spacesuits
NASA’s Artemis program promises to land the first astronaut on lunar soil by 2024 and, to do so, it will need several new instruments and technologies. Among the needs is a new generation of spacesuits, more comfortable and safer, that will allow the astronauts of the next missions to work efficiently on the Moon.
This new generation of space suits has been designed with the help of artificial intelligence, especially for life support capabilities. These tools are capable of regulating temperature, oxygen level, and pressure so that the human body functions and survives in the inhospitable environment of space and the lunar surface.
Among the series of tasks of the AI in the design are processes of calculation and design study to combine the highest level of safety and robustness of these types of equipment in the lowest possible volume and weight. Those responsible for the project indicate that the approach is much faster, generating new versions and concepts at an impossible speed for a team of engineers.
4. Improvement of Images and Videos of the Apollo Program
The Apollo program, which took a man to the moon in the 1960s, condensed a large volume of photos and videos. Although the photos were high resolution, the video technology of the time was very limited and this compromised the quality of the images recorded in the six missions that reached the lunar ground.
One of the initiatives can be seen on the DutchSteamMachine channel on YouTube, which specializes in restoring historical videos with artificial intelligence. The enhancement is done by cleaning artifacts from the images, and increasing resolution, and color for the black and white originals. Also, there is an increase in frame rate per second, making the captures more natural.
5. Navigation in Martian Rovers
NASA has rovers on Martian soil. They are the Curiosity, in operation on Mars since 2012, and the Perseverance, currently on its way to the red planet. The exploration vehicles have navigation systems that use AI to prevent the robots from colliding with obstacles, falling into holes, and ending up trapped in some accident on the surface.
As the distance between Mars and Earth varies, a signal from NASA controllers takes different times to reach the destination. The operation can last a few minutes when the two planets are closer, or up to 24 minutes, at the furthest point from their orbits. In both cases, it is a long time for a response to leave the Earth and prevent an accident.
In this sense, the presence of AI gives a certain level of autonomy to the equipment and acts faster, preventing accidents that could interrupt the research. The success of this technology can be observed in the resilience of Curiosity, in operation on the planet for eight years.
6. CIMON: A Robot in the International Space Station
CIMON 2 is the second generation of a robot with artificial intelligence sent to the International Space Station (ISS). Part of a program of the German space agency (DLR) and ESA (European Space Agency) in partnership with Airbus, CIMON was designed to be a companion for astronauts in the space lab.
To be able to evaluate the astronauts’ stress levels during interactions and conversations in the middle of the routine at ISS, CIMON 2 relies on technologies developed by IBM. The idea of the program is to study ways in which humans and intelligent machines can be used in synergy in space exploration.
An example used by the creators of the project is the idea that on a trip to Mars, astronauts would be isolated from humanity for up to two years. Having access to a robot astronaut is not only a form of distraction but can also prove to be a relevant technical resource in solving problems that can appear from one hour to another throughout missions in space.