HVM Catapult has launched a dedicated AI model center to provide independent evaluation and validation of artificial intelligence systems within the manufacturing sector.

HVM Catapult has launched a dedicated AI model center to provide independent evaluation and validation of artificial intelligence systems within the manufacturing sector. Led by Professor Chris Dungey, the center focuses on ensuring that AI-driven digital manufacturing solutions meet rigorous safety, security, and regulatory standards before industrial deployment. This initiative operates across the HVM Catapult network of seven centers in the United Kingdom, aiming to bridge the gap between experimental AI research and reliable, production-ready manufacturing workflows.

The manufacturing industry currently faces a significant barrier in adopting AI due to the lack of standardized testing frameworks for algorithmic reliability in safety-critical environments. While software vendors often provide performance metrics, independent validation is essential for sectors like aerospace and medical device manufacturing where failure is not an option. By positioning itself as a neutral arbiter, HVM Catapult addresses the trust deficit that slows the integration of AI into complex processes like LPBF or DED monitoring. This move aligns with broader global trends toward AI governance, providing a necessary service layer for companies struggling to verify the robustness of their digital manufacturing stacks.

Manufacturers should view this center as a resource for de-risking the integration of AI into existing quality control and process optimization pipelines. The practical value lies in the center's ability to provide objective benchmarks for AI models against specific industrial use cases, reducing the liability associated with black-box algorithms. Success for this initiative will depend on the center's ability to establish standardized testing protocols that are recognized by regulatory bodies and industrial stakeholders alike.