SEAPROP's general objective was to create a predictive analytics system to support the planning and optimization of industrial production, based on the various data available on the production processes of the sponsoring company, both those captured in real time by its systems and the historical data on the processes themselves, orders and invoicing.
From the historical order and invoicing data available in the enterprise resource planning (ERP) system, future orders can be predicted, and from production data the manufacturing time of a given product can be estimated. However, since manual adjustments to the machinery during each process can cause faults and alarms that alter execution times, the repeatability of the processes is not guaranteed, which limits the accuracy of intelligent data analysis techniques based solely on deterministic models.
To deal with this lack of repeatability, the project turned to stochastic modeling and analysis of the manufacturing processes, complemented by optimization algorithms capable of determining the work sequence that would make the most of the available human and material resources while meeting delivery deadlines. CTIC collaborated with Técnica de Conexiones, S.A. (TEKOX) on the design and development of the demand forecasting algorithms, the stochastic simulation models and the optimization algorithms, as well as on the functional testing of the resulting system.
INNOVA-IDEPA Program (RIS3-Business Program), 2017 call
