Mapping & matching – the allocation and linking of master and variable data from various sources – is essential for being able to analyze and visualize the material flow across the different sites, facilities and production steps based on a difficult source of raw data. The customer had already started to manually map its data. ONE LOGIC paved the way for this process to be automated via machine learning and for the introduction of highly complex data management. To achieve this, a proof of concept was drawn up using the Data Product Platform ONE DATA. The data was compared, aligned, analyzed and – in part – reclassified using machine-based pattern analysis according to the same entities, including volume, weight, frequency, duration of storage, etc. This enabled ONE DATA to predict that material A and material B were identical and could be used for the same order. Visualization in ONE DATA allowed the material flow to be tracked and made transparent for the first time across the different sites and production steps. A specially created cockpit was used to further optimize the process. Pattern matching algorithms, decision trees and sequence mining were used, for example, to identify and visualize delivery delays and possible influencing factors. In addition, newly developed algorithms in ONE DATA meant that the entire process could be safeguarded.
As an end-to-end solution, ONE DATA can be used to analyze even complex material flows and visualize them. The integration and linking of data is much quicker and more accurate, paving the way for reduced inventories and faster delivery times. The central Data Product Platform ONE DATA can also be used for other use cases, such as the AI-based analysis of customer relations management.