Institute of Robotics and Information and Communication Technologies of the University of Valencia (IRTIC) participated in an initiative funded by the European Space Agency (ESA), which aims to help successfully carry out missions that send robotic rovers – or other autonomous equipment – into space. Specifically, the goal was to provide a solution and respond to the lack of data that spatial artificial intelligence applications have.
Sector AI together with Augmented Reality (RX) They have grown very quickly and, for example, are used to make useful models in the medical sector. In view of this, IRIC points out that the aviation sector has also attracted interest; actually, ESA has promoted the integration of AI into the data systems of its missions and developed a roadmap focused on automation (A2I) and in its Agenda 2025 dedicated a leading place to AI and RX.
This is where the initiative comes from. «AI-Aided-XR: AI Aided eXtended Reality applications», which was developed from October 2023 to June 2024 and focused on training and testing autonomous AI control devices on the lunar surface for which, among other thingscreates simulations of the lunar soil and surface, contributing to space exploration. It is relevant because it should be noted that one of the problems in this sector is the lack of data and photographs of some areas of the Moon – for example, the far side, where China recently arrived – or the surface of Mars.
Valencia team IRTIC who participates in the project, has created software that virtually recreates surfaces, so simply put, rovers that are sent on missions into space can “train” earlier on Earth, on surfaces similar to those that can be found.
“Interestingly, samples with the same pattern can be generated so they can serve as training for artificial intelligence algorithms for autonomous navigation, such as landing well on the moon. Rovers need a lot of data to train on, and if no real data exists, it must be created. We create software and 3D simulations to generate rover training sets»Indika MFernández Marín arches, director of IRTIC’s ARTEC group. It involves combining virtual reality and artificial intelligence to synthetically create datasets in a simulator.
So they are “synthetic images, not of the real moon”, although no one would tell from the looks of it because they look like photographs. “It was a project to explore the technology if it could work, and it was seen that it could, and the idea is to generalize it,” say IRTIC.
Previous experience on Mars
Five experts from IRTIC were involved, although there are a total of 25 in the laboratory. The team was already a specialist in terrain generation, Because they had previously created a representation of Mars for training and dissemination, that’s why the company that commissioned the project contacted them, says Fernández.
As the institute explained, the international community can provide these recreations of lunar soil or other celestial bodies using AI and RX that include educational facilities that use augmented reality to interact with satellite imagery; or for commercial tourism; and mission operators may be able to view automatically tagged and classified terrain images taken by planetary rovers.
AI also plays a key role in research opportunities arising from the Mars Sample Return (MSR) mission due to the importance of computer vision techniques in estimating the location of objects.
An initiative funded by the European Space Agency (ESA) It was developed from October 2023 to June 2024 and has a stake in IRTIC through the ARTEC group. It forms a project consortium with GMV NSL Ltd from the United Kingdom, GMV Soluciones Globales Internet SAU from Spain, GMV Innovating Solutions Sp from Poland and ESA European Space Operations Center from Germany.
The result of the year-long work was presented at the “9th European Mission Operations Data System Architecture Workshop (ESAW)” organized by ESA on June 13 and 14 in Germany. The event has been held since 2005 and in its most recent editions has already hosted such innovative and recent fields as big data, augmented and virtual reality, machine learning and quantum computing.
