NASA is using AI to make hardware that’s “three times more powerful.”

Space agency NASA has begun using artificial intelligence to design its mission hardware, designing components that it believes are significantly stronger than their human-designed counterparts while shaving two-thirds the weight.

Developed by research engineer Ryan McClelland, the Evolved Structures process takes a fraction of the time required by NASA’s experienced designers and relies on a generative algorithm to create metal supports and brackets for various space exploration missions.

“People, maybe they do an iteration every week or two if things are going well,” McClelland explained on NASA’s Small Steps, Giant Leaps podcast.

“The AI ​​does something on the order of one iteration per minute. So you get a lot more iteration cycles, and because of the more iteration cycles, you just get more optimal designs much, much faster.”

NASA uses AI to design components for its EXCITE telescope (top) and spectrometer (top).

So far, the system has been used to design everything from a truss for NASA’s balloon-borne EXCITE telescope to an optical bench for an ultraviolet imaging spectrometer to house its optical components.

“Of the current applications, the optical bench is probably the most impressive,” McClelland told Dezeen.

“It’s a radical departure from typical optical benches and has far better structural performance. It also condensed what would have been around 10 parts into a single part that can still be CNC machined.”

Ryan McClelland shows a structural mount for the Survey and Time-domain Astrophysical Research Explorer (STAR-X) mission.
Engineer Ryan McClelland developed the generative design process

Much like the chatbot ChatGPT or the image generator DALL-E, the system still relies on human input in the form of a precise briefing detailing the requirements of the part, including the load it must support and the forces it will be subjected to will be.

This data is fed into the generative design software, which can create 30 to 40 iterations in a few hours, each improving the last one to develop an optimal structure.

“The AI ​​creates the design and then tests the design through finite element analysis to make sure it works to verify requirements and then also runs manufacturing simulation to make sure it can be manufactured,” explained McClelland on the podcast.

This means that the final design can be fed directly into a digital manufacturing process and machined from the CAD model by a standard CNC router.

From design to production, this process can take as little as a week. McClelland estimates that this is about 10 times faster than NASA’s normal process, in which the design is passed around between a designer, a stress analyst who verifies its performance, and a machinist who tests whether it can be manufactured.

“What the Evolved Structures process does is the back and forth that goes on between a bunch of different people – and that can take months or years depending on the project and how committed the people are and if they’re working on other things – and it reduces that to something that’s all done by the computer,” he said.

Ryan McClelland holds an AI designed component for astronomical instruments
The parts are manufactured using a conventional CNC machine

The resulting components feature “almost bone-like” organic shapes that are able to withstand higher structural loads than human-made parts.

In fact, McClelland found that the AI-designed components had up to 10 times less stress concentration while saving up to two-thirds of the weight.

“The structures tend to work much better,” he said. “They’re somewhere on the order of three times better in performance.”

Ryan McClelland looks at NASA parts
McClelland believes the system could help NASA save time and money

Given that NASA manufactures thousands of custom parts for its various missions each year, McClelland predicts that the design process will become common practice when designing structural parts, electronics, and other subsystems in NASA’s instruments and spacecraft.

This, in turn, would help reduce both the time and costs associated with space exploration.

“The space station holds six or seven people, but it costs $100 billion,” he explained. “I really believe that AI has the potential to drastically reduce the cost of developing these complex systems because it’s really great at things like that.”

Previously, German software company Hyperganic used AI to develop a prototype rocket engine that was 3D printed in one piece.

The photograph is by Henry Dennis.

Leave a Reply

Your email address will not be published. Required fields are marked *