TRENT. The collaboration between the University of Trento and ESA – the European Space Agency is strengthened. For the next three years, the Department of Industrial Engineering (Dii) will try to understand how artificial intelligence can support space missions.
The principal investigator of the project is Paolo Rech, Dii researcher who has been working on the reliability of processors for terrestrial applications and now also for space research for over fifteen years. Together with his team, Rech will develop automated systems, i.e neural networks dedicated for use in orbit, that can support the work of astronauts, facilitating them for example in operations that today they perform outside the spacecraft, or that can recover useful information from the observation of the planets.
But not only. An enormous amount of data arrives from satellites, probes or planets which, once on earth, must be processed and interpreted. A process that requires considerable bandwidth and very long times. Just consider that it takes several minutes to send and receive messages between Earth and Mars. Whatever movement it needs to make, the space rover must therefore wait for this delay before executing it. This is an extremely long time, which makes it impossible to timely make the urgent and rapid decisions that arise from real-time data processing. «This is why it would be important to put accelerators on board satellites. To make the actions of robots and the decisions of space probes automatable”, explains Rech.
