A new model of artificial intelligence could speed up the diagnosis of physical malfunctions in spacecraft and space flight systems. Scientists are confident that this will improve mission efficiency by reducing downtime.
Artificial Intelligence Research for Spacecraft Resilience (RAISR) is software developed by Pathways trainee Eva Gizzi, who works at NASA’s Goddard Space Flight Center, Maryland. With the help of RAISR, the model can diagnose malfunctions in spacecraft and space flight systems in general in real time.
“A spacecraft that reports a malfunction is like a car with a Check Engine light on,” the researchers note. – You know that there is a problem, but you cannot pinpoint or find the cause. This will help the RAISR algorithm, which diagnoses the exact cause. ”
The scientists noted that the current diagnosis of faults depends on how simple the mechanics of the device are and how well studied by the researchers. For example, if the temperature of the instrument drops too low, the spacecraft can detect this and turn on the heaters. If the line current jumps, the spacecraft can isolate the faulty circuit. However, if we are not talking about cases where the device knows about a malfunction and understands ways to solve the problem, then scientists need more advanced models.
Such conclusions require the ability to follow a logical chain of non-trivial inferences – a kind of human reasoning. Artificial intelligence (AI) will cope with this, which knows how to associate a decrease in the temperature of a spacecraft with a malfunction of its internal heat regulation system.
Transmitting such faults to humans on earth not only takes time, but also requires valuable resources – communication networks and bandwidth for small missions in near-earth orbit or even for exploring distant planets, where the bandwidth of communication channels with dispatchers on Earth is limited by distance.