REF 2021 Impact Case Study: Improved anomaly detection algorithms to benefit manufacturing and aviation

Impact

Description of impact

University of Portsmouth researchers have developed anomaly detection algorithms which are better at detecting faults, revolutionising two different sectors’ engineering practices. Our algorithms were deployed in three dairies operated by Stork (a supplier of dairy filling machines), where 31 faults were detected early. These faults would have led to chains of escalating damage and cost (breaches of contract, lost product, equipment repair bills) costing GBP15,000,000. These algorithms have also transformed how flight data is analysed for a civil aviation company (Flight Data Services) resulting in the development of a world first product (Express Readout) to detect errors in flight data recorders. This key product led to the company’s acquisition by a multi-billion dollar defence firm for GBP8,000,000 including the purchase of the IP that linked directly to our research.
Impact statusClosed
Impact date20132020

REF

  • REF2021