2024: Testing of improving supply reliability at Saku Metall Allhanke Tehas through early warning system

Experimenting in Improving supply reliability at Saku Metall Allhanke Tehas AS through early warning system. 

In this demonstration project, we addressed a key challenge related to processing orders from our largest client. The main objective was to identify orders that require immediate attention to specific critical parts, materials, or purchased components—as delays in their procurement and logistics could compromise supply reliability and disrupt production timelines. 

To solve this, a customized AI tool that leverages the strengths of Association Rule Mining (ARM), while being extensively tailored to the structure and complexity of the dataset provided by Saku Metall was validated. 

The AI tool integrates standard data mining metrics, such as Support and Confidence (commonly used in the Apriori algorithm), with a novel tokenization strategy designed specifically for analyzing multi-level Bill of Materials (BOM). This approach enables the identification of rare purchases and unique behavioral patterns within the BOM data.