To path indicator Subpages of “About Us“ To navigation To quick access To footer with other services
To content
Logo IPS
Faculty of mechanical engineering
The picture shows the old building of the Faculty of Mechanical Engineering. © Jürgen Huhn​/​TU Dortmund

Successful kick-off workshop of the PrABCast research project

Foto einer Reihe von Autotüren, die vertikal in einer Lagerhalle oder Produktionsstätte aufgestellt sind. Die Türen sind in verschiedenen Farben, darunter Schwarz, Blau und Rot, und sind dicht an dicht angeordnet. Metallstangen halten die Türen in Position. Der Fokus liegt auf den Türen, die von vorne bis hinten in einer Linie zu sehen sind, wodurch eine Tiefe im Bild entsteht. Die Szene vermittelt den Eindruck von geordneter Lagerung oder Vorbereitung für den weiteren Einsatz in der Automobilproduktion oder -reparatur. © Pixabay
At the kick-off meeting of our new research project PrABCast, carried out by the partner institute RIF e.V., the project partners had a lively discussion about the demands on companies' sales and demand planning. In addition, possibilities were discussed to increase the flexibility of customer order-oriented companies with regard to the fulfilment of individual orders and to open up data sources that enable better forecasts.

Companies with an early customer decoupling point in production are faced with the conflicting goals of ensuring a high degree of flexibility while at the same time fulfilling individualised customer orders.

In order to discuss this area, the PrABCast research project started with a meeting of the project-accompanying committee. ERCO and RapidMiner were able to introduce the topic with a presentation on a joint project in sales and demand planning. Andreas Bohlmann (ERCO) reported on the requirements his company has for sales forecasts. RapidMiner gave insights into their part of the work in this project, where they improved ERCO's forecasts using machine learning.

In the following workshop, the requirements of the participating companies for their sales and demand forecasts and for the use of machine learning were collected and discussed. Possible sources for data enrichment, which currently still have an influence on the forecasts, were also sought and found. As a result, it can be said that all participants assume that a good forecast of demand, supported by AI, will lead to an increase in flexibility in responding to customer wishes.

The success of the workshop is not only due to the different perspectives of the users and system providers, but also to a broadly diversified requirements profile from the different industries of the respective users.

In the next steps, the use cases are to be specified by the project partners. In this process, new data sources are to be identified and forecasting models integrated. In particular, the inclusion of expert opinions, which have been decisive in most companies up to now, seems promising and holds potential.

Back

To top of page