That would reduce the load on manual monitoring by teams of experts.

Thousands of satellites are being launched every month by companies such as SpaceX, OneWeb, and Amazon as they jockey for space supremacy and build mega-constellations that orbit the earth. Current launches, in a month, exceed more than the annual figure just a few years ago. As a result, there is a growing risk of collisions among spacecraft and debris in the orbital space around the Earth. The European Space Agency is pressing AI to automate collision avoidance procedures. (Space.com)

SpaceX had a near miss last month

Last month, the astronauts of SpaceX’s Crew-2 mission for NASA, ensconced in their Crew Dragon capsule, had to scramble to take precautions because a piece of space junk passed dangerously close to their spacecraft.

According to the ESA, there are currently 128 million objects larger than a millimeter in orbit today, and these present a growing concern for both astronauts and satellites.

The ESA regularly holds conferences on tackling the menace of space debris. It revealed at the 8th European Space Debris Conference held virtually from Darmstadt Germany, April 20 to 23, that teams of spacecraft controllers have to conduct avoidance maneuvers to protect their low earth satellites from space debris.

These manual processes burn up fuel, are expensive to conduct, and often interfere with the collection of important scientific data.

AI to deal with space junk

The ESA has requested help from the global AI community to develop an early warning and dodging system that would operate autonomously against debris.

Working with a large data set of past conjunction warnings, AI experts predicted the evolution of collision risk of each alert over the three days following the alert.

“The results are not yet perfect, but in many cases, AI was able to replicate the decision process and correctly identify in which cases we had to conduct the collision avoidance maneuver,” said Rolf Densing, Director of ESA Operations.

However, the results are useful to ground-based teams as they evaluate and monitor each conjunction alert, by warning if a satellite is at risk of collision.

According to Tim Flohrer, the Head of ESA’s Space Debris Office, who was speaking to Space.com, newer approaches to AI development, such as deep learning and neural networks, would be required to improve the accuracy of the system.

The ultimate objective, according to Densing, is to set up a collision-avoidance system similar to modern air-traffic management. It should work completely autonomously without human intervention.

Related Story:  The European Space Agency Will Use AI to Burn Space Debris

Image credit: Wikimedia