How is AI Accelerating the Race to Get Rockets in Space?

AI is reshaping how space agencies and aerospace manufacturers approach rocket construction, moving the industry beyond conventional production methods towards automated, data-driven processes that could fundamentally alter spacecraft development.

The European Space Agency, working alongside German aerospace manufacturer MT Aerospace, has implemented machine learning across three critical manufacturing processes:

According to ESA, these AI-driven initiatives are already demonstrating significant advantages in materials processing and production efficiency.

MT Aerospace, an international aerospace company employing more than 500 people, is applying these new techniques to materials processing. The focus areas include shot peen forming, friction stir welding and carbon fibre placement methods – all traditional processes now enhanced through artificial intelligence.

Machine learning reshapes metal forming

Shot peen forming represents one area where AI is making substantial inroads. This process involves shooting metal with small balls to bend it into shape without heating, ensuring the resulting metal remains strong and resistant to fatigue.

MT Aerospace uses this method to shape the dome heads of Ariane 6 rocket fuel tanks.

The unpredictable nature of high-speed ball impacts has historically made precise outcomes challenging. However, machine learning is now being used to predict how metal will deform, providing what ESA describes as a fast and precise method to reach desired shapes with just 2 mm tolerance.

It is the first time such predictive technology has been applied to this manufacturing process, according to ESA.

AI accelerates welding analysis

Friction stir welding, which has replaced traditional arc welding in space applications, represents another area benefiting from AI.

This technique heats metals by rotating a pin over the welding area at high speeds, using friction to fuse materials together for stronger structures like Ariane 6 tanks.

Machine learning is now helping set up machines faster, supporting documentation efforts and automatically checking final weld shapes. According to ESA, the automatic evaluation of weld seams has reduced analysis time by 95% compared to traditional processes.

"Artificial intelligence, such as machine learning, in combination with new digital technologies is transforming launcher manufacturing," says Daniel Chipping, ESA project manager for software-centred and digitalisation activities at the Future Launchers Preparatory Programme in Space Transportation.

Daniel adds: "From automating complex analysis tasks to reducing tedious machine stop-starts, we are starting to see the benefits across all materials and shaping processes."

Defect detection in composite materials

Carbon-fibre reinforced plastic offers shapes that are lighter and stronger than traditional materials.

The Phoebus project is exploring the use of carbon-fibre tanks for Ariane 6, built in layers to optimise strength and weight.

MT Aerospace is integrating new laser sensor technology that, powered by machine learning models, will detect and classify defects automatically. This capability could keep production lines running while shortening overall production times.

The work stems from ESA's Future Launchers Preparatory Programme, which is investigating AI applications in materials processing across the space industry. The programme is exploring how artificial intelligence can develop better processes and entirely new shapes in materials for future rockets and spacecraft.

MT Aerospace brings expertise in additive manufacturing, metalworking, carbon-fibre reinforced plastic and hydrogen technology to these collaborations. The company develops, manufactures and tests components for institutional and commercial launch vehicle programmes, aircraft, satellites and applications in automotive and defence industries.