UK DiMOS project funded by Innovate UK reaches key milestone
02/06/2020
The Innovate UK-funded Digital Monitoring of Ships (DiMOS) project is now actively collecting data from a number of critical installations, both on land and at sea.The DiMOS project bridges the gap between conventional learning artificial intelligence (AI) systems and intelligent expert-led condition monitoring systems by building the typical machinery failure profile knowledge and the local risk needs into the system at start-up. This is then optimised by gaining the local detailed knowledge of operational behaviour during the early commissioning phase.
Innovate UK, a government funding and support group, drives productivity and economic growth by supporting businesses to develop and realise the potential of new ideas, including those from the UK’s world-class research base, and recognises the benefits to society by leading developments in this domain. As a result, around £1.5 million has been assigned to the completion of this initiative.
DiMOS can report that, given the unusual circumstances that prevail at this time, it is on target and now receiving data. As a result, it is now in development of the prescriptive maintenance planning platform tools for machinery systems. This will be achieved via the use of real-time sensor data and AI-based models to prescribe maintenance based on the monitored condition. This is derived from machinery reliability templates supported by local failure mode effects and criticality analysis (FMECA) data and the detailed minutiae of data flowing into the model architecture. The risk level, scheduled maintenance and timing and cost management at the asset level are therefore considered.
DiMOS itself is formed from a number of expert UK stakeholders, namely Vibtek Ltd, CMServices Global Ltd, TWI and Brunel University, and represents a true partnership between industry, academic research and government initiative. This project is now part way through its scheduled two years and is on target to be completed by the end of 2020.
DiMOS has been particularly successful in securing industry stakeholders who recognise the need to reduce vulnerability by gathering intelligence without the need for additional, or the diversion of existing, staffing. The need to work remotely has come sharply into focus as a result of the current COVID-19 crisis, where such benefit is real. Working with industry, offshore and marine operators, DiMOS is now collecting data periodically and in real time from a wide number of assets considered typical in these industries, such as pumps, fans, compressors, gearboxes and hydraulic systems. In the main, it is collecting vibration data via installed sensors and data collection routes, but is also incorporating process, environmental and other key data fields that can be used to provide a complete picture of condition, a diagnostic and prognostic capability. Using feature extraction techniques, DiMOS is building a platform that can look at the component, its machine and the parent plant, to assess condition, any deviations from what is expected and the significance of any deviation to deliver the most suitable advisory output. Coupled with the combined experience of the diagnostic standard and the signatures of typical failure modes, the DiMOS deliverable will provide a balanced yet specific prescriptive advisory output specifically designed to be actioned by the maintenance crew as part of its normal duties.
The real benefits of this type of platform are derived from the elimination of measurably unnecessary tasks, focusing engineering resources on tasks that actually add value and improve the through-life cost of asset ownership, while also adding the early warning capabilities of impending machinery failure surveillance.
This is an ambitious project that will result in a set of tools that can be deployed to deliver a meaningful and real step change in the maintenance and reliability of critical machinery.
