Assessing AI Feasibility across Vertical Mill Operations
Abraham Moon & Sons is a mill that needs no introduction – founded in 1837 and one of the last remaining vertical woollen mills in Great Britain, it is no overstatement to say that the history of textile manufacturing in the region is very much interwoven with the history of the mill. In its near 200 years of operations, Moon has developed a reputation for producing fabrics of the utmost quality and cutting-edge design. To stay ahead of the curve when it comes to textile manufacturing innovation, Moon set out to work with Future Fashion Factory, scoping the feasibility of Artificial Intelligence software across all areas of operations within a textile mill.
About the Project
Working with Professor Ningtao Mao at the University of Leeds, Moon and Future Fashion Factory projected the usability of software that would allow the user to define performance requirements, whether this be machine efficiency, customer service levels or a blend of both. Using the results of these studies, the mill would then be able to create a production schedule with critical path timings that could result in achievement of the performance requirements. The software could then potentially ideally integrate into existing enterprise resource planning software so that live data is used and manual input is limited.
It stands true that the variety of ways in which AI could be implemented within a fully vertical mill is a reflection of the seemingly infinite number of permutations that a business of Moon’s scale can have when planning the production of its own raw materials and processes.
Traditional software used in mill operational management does not typically have the ability to schedule production, relying instead on a variety of spreadsheets and reports to make decisions around what to produce, by when and in what volumes. This in turn carries an obvious risk when it comes to minimising efficiency and waste.
Advanced Planning and Scheduling Tools
Following a full-scale investigation into the many pathways going forward in which AI could streamline Moon’s operations, the Call 3 project was able to take its findings and help inform decisions around a new production planning system. While AI certainly carries the potential to produce an optimise production schedule, the project’s findings directed Moon down an alternate route – to introduce SAP Advanced Planning and Scheduling software (APS) as a standard feature.
APS is especially well-suited to environments where simpler planning methods cannot adequately address complex trade-offs between competing priorities – this can certainly be said of a vertical mill. As opposed to AI’s use of machine learning, Advanced Planning Systems uses complex mathematical algorithms to achieve its results: to scope demand, to plan and schedule production within specified constraints, and to derive optimal source and product-mix solutions. Notable to APS in the context of production scheduling is a ‘heuristic approach’ or rule-of-thumb method, nevertheless sufficient for reaching an immediate, short-term goal or approximation – and critically, without requiring the huge data-sets as needed for machine learning. By the same token, the increased adoption of Artificial and Machine learning has only benefited APS systems, where new algorithms have been adopted to solve even more complex planning problems.
Following the implementation of these findings, Future Fashion Factory looks forward to seeing how Moon continues to lead the way as a vertical mill, and is thrilled to have helped inform their decision to embrace APS solutions as part of our collaborative Call 3 project.