In many industries efficiency is severely constrained by quality assurance issues arising from complex manufacturing process, where faults commonly do not arise from a single point of failure, but are a consequence of a number of contributing factors. Assessing the effect of these factors and applying remedial actions may require expertise from a range of different disciplines and experiences, which for operational and logistical reasons may not be available at every stage in the manufacturing process. The cost of non-conformance to quality standards is more than just the simple issue of a devalued product being produced. Lack of predictability and unexpected variation in quality can severely influence the delivery reliability. For exampe, during the planning process an estimate is included of the proportion of the goods manufactured which will conform to required standards (Product Through Yield). If the estimate is too low then there is wastage in the form of rework which adds cost and affects speed of working, reapplication to lower margin outlets or scrappage, if the estimate is too high then this means that replacement-manufacture has to take place right back at the beginning of the process, which could entail significant delays, or the whole reschedule of the plant.

The Project

This project directly addresses the challenge for organisations to realise actionable knowledge from an ever increasing flood of potentially valuable data. Currently, the need for skilled ‘data scientists’ is a major bottleneck in this regard. The project will apply existing and develop new technologies to create an ICT tool set which alleviates this bottleneck through provision of a collaborative platform with tools for data integration and analytics deployment which are accessible to non-ICT specialists. The project includes innovative elements especially concerning the integration of text-based knowledge with numerical data and inclusion of ‘plug-in’ algorithms for model-based novelty detection in data streams. In this way the envisaged toolset provides a vehicle for projects to move from problem definition, to knowledge discovery and on to real-time deployment, and with an embedded continuous improvement cycle. Each instance of the toolset will also become a knowledge system in its own right which can be interrogated by users, either directly in relation to the process route and quality issue served, or as part of the wider organisational knowledge base. A means is therefore provided for non-specialists to organise and progress data analytics projects independently, but where skilled data scientists are available, for them to use the framework to structure and steer project work in multiple areas simultaneously, thus increasing effectiveness.
The project is funded by Innovate UK, start date 1.10.2014 and has a length of 18 months.