Sequence planning in manufacturing has been done manually up to now, and the complexity of taking all the framework conditions
into account is constantly increasing.
Research and development of an AI application for optimised sequencing using
product and production data.
The usable output is optimised and thus the added value is increased by saving time in the manufacturing process.
Air conditioning and ventilation technology, components, devices and system technology: Westaflex is a German family business that has been providing architectural building services and air ducts since 1933. Solutions for a better climate and simplified construction made of aluminium, stainless steel and premium plastics. For every type of building and pipe, there are ventilation systems in many shapes and materials, in different diameters and grooves, so that the systems can be flexibly combined and adapted to the requirements. Accordingly, many different set-up processes are required in production, which in turn requires precise planning. To ensure that everything runs smoothly and as efficiently as possible, the sequence of orders must be planned precisely before production.
Sequence planning determines the production sequence of individual orders with the aim of making the production process lean, time-saving and thus efficient. Currently, the sequence planning of orders at westaflex is still done manually, for example using spreadsheets or analogue planning boards, and not automatically. Factors such as machining, set-up and delivery times, product specifications, material or the reduction of effort must be taken into account individually. This is very cumbersome to implement: Orders and sequences of machines cannot usually be simply shifted without having to manually adjust the rest of the planning.
Data analysis and AI make for intelligent production planning
The goal is therefore to optimise the sequence planning of production orders with the help of artificial intelligence. For this purpose, a wide variety of data, for example ERP data and real-time data from production, are being evaluated in order to derive indications for optimal machine occupancy and to use these findings for sequence planning.
For this purpose, a data platform is being developed as a web-independent on-premise solution. Order data, resource data, process data, tool/maintenance data, logistical data and monetary data are brought together on the data platform and processed for the AI application. The data platform thus represents an IT infrastructure equipped with interfaces for the plant’s internal data and the data from the AI application. The solution found is implemented and tested with real data, optimised and validated.
For the AI Marketplace, this project results in valuable solutions and building blocks in the field of production system development, which also benefit other companies facing similar problems or goals. Westaflex primarily expects this project to optimise the usable output and thus increase
the added value by saving time.