Initial Target Type 45 – Potential Pull-through to Type 26
Britain’s Ministry of Defence (MoD), through its Defence Innovation Fund and the Defence and Security Accelerator (DASA) programme, is funding the development of a progressive, unified asset health management and risk forecasting tool using state of the art data science and analytical modelling, which could potentially save front line forces an estimated significant savings if successful
The two-stage recurrent neural network, being developed by decisionLab and sponsored by Joint Forces Command, is part of DASA’s fast track ‘Revolutionise the human information relationship for Defence’ competition. The product was originally developed for civil aviation applications, but the funding is helping to repurpose it for the military.
The hope is that it will in time provide an insight into the future, allowing maintenance engineers to view the status of systems and their predicted health a day, a week, or even a fortnight in advance. Gripped by this opportunity, the Royal Navy has invested £150,000 in the development of the neural network for exploitation on-board a Type 45 destroyer, and pull through onto the Type 26 – if proven successful.
A Royal Navy ship is incredibly complex, and the Type 45’s systems can record 10 million data points a day. With such a large, complex dataset, the type of machine learning offered by the neural network will likely have a significant impact on maintenance schedules and support, improving capability, saving money and delivering efficiency.
At present, decisionLab is training its neural network on 1.8 billion lines of Type 45 Platform Management System data. Each day the system gets smarter and more capable, and under current development plans this system will be installed onboard HMS DIAMOND for a trial this summer. It will allow the user to validate system assumptions and help contextualise events to further train and improve the model.
“The Defence and Security Accelerator competitions provide the Royal Navy with a unique opportunity to both engage with a broad spectrum of small to medium sized enterprises, whom are often new to the Defence market, and to grip technological opportunities from disruptive markets and apply them to Defence problems. This rapid development process, with collaboration at its core, will provide battle-winning capabilities to the hands of the user,” commented RN Innovation Programme Manager, Lee Packer.
“This is a clear demonstration of the cultural shift across all organisations to focus on capability integration and exploitation as well as technology development. This project is a great example of collaboration between DASA, the competition sponsors (Joint Force Command) and the front lines working hard to turn technology into true capability [….] The Navy’s commitment to integrate the decisionLabs project onto existing architecture will help provide an accurate value proposition for future procurement across the maritime capability and possibly beyond,” added DASA Exploitation Lead, Joe Hemming.
The ‘Revolutionise the human information relationship for Defence’ competition is the first competition launched since the start of DASA in December 2016: some 34 projects have been awarded funding, with decisionLab being one of seven to receive phase 1 fast track funding under Challenge 2. In October 2017, decisionLab was one of two companies to successfully progress to Phase 2 to continue development of its neural network system.