It is a challenge for small and medium-sized wholesale companies to assess the sales development. Standard retailing systems provide forecasts but are not able to calculate possible deviations. Together with its project partners, FIS would like to change this by using artificial intelligence (AI) for the OBER project (optimum inventory planning for resource conservation).
Controlling warehouse stocks in the best possible way means to avoid both delivery bottlenecks and excessive stockpiling. The currently used sales forecast models only provide point forecasts. They cannot predict the probability of the difference between forecast sales and actual sales. MRP controllers have to decide themselves to what extent they rely on the forecast for their planning. Consequently, many companies hold unnecessarily high safety stocks.
AI might improve this situation. Together with its project partners Fraunhofer-Arbeitsgruppe SCS, Trevisto AG and Eisen-Fischer GmbH, FIS develops and tests an AI procedure which can be used to determine the optimum warehouse stock. Aspects such as quantity-based purchase conditions, logistical lot sizes, storage costs etc. are also considered.
Within the scope of the OBER project, Eisen-Fischer therefore implements the new forecast tool in parallel with the methods previously used. In this way, AI forecasts, conventionally created forecasts and actual market developments can continuously be compared over the entire project duration of 32 months. From the results, AI will “learn” to make more precise forecasts. The aim is to develop a user-friendly software solution, particularly for SMEs.
As a result, the connection to SAP implemented by FIS is not only used to integrate all SAP data, but it also ensures that the new procedure can be integrated into existing SAP systems without any problems.
Read here more about AI & Machine Learning with SAP.
