Skip to Content
Author's profile photo Timothy Schroeder

Mining your own supply chain data with machine learning

We are all hearing more and more about Machine Learning. Sometimes I have found it difficult to discern the difference between machine learning and other types of artificial intelligence. I sought out information from experts and here is what I learned. Machine learning experts have devised a series of questions to ask yourself about the use case you are considering. Do you need automation? Can you formulate the problem? Do you have the data? And more. Machine learning experts in collaboration with business process experts can answer these questions to validate use cases. I also learned that high volume tasks with complex rules and unstructured data are good candidates. Lastly, does the problem have a regular pattern? “Machine learning learns regularities and patterns.”

Great, now I know how use cases are properly validated.

This better understanding got me thinking about information that resides in ERP systems which are a challenge for users to maintain. What if machine learning could help maintain master data values related to supply chain processes? This is interesting topic because it can help improve the supply chain information wholesale distributors act upon in our system thousands of times a week such as:

Planning and purchasing information

  • Supplier delivery lead times
  • Stock-in – transit lead times
  • Safety stock
  • Returns reason codes trends

or information wholesale distributors act on monthly or periodically to improve supply chain planning, such as:

  • Stock trending toward slow mover
  • Moving a material status from unplanned to planned
  • Sales order stock turnover

These all are tedious tasks for users, and the data resides in your ERP system.

Is it time for you to start thinking about machine learning? Do what I did. Learn more about machine learning and how it is properly applied to data. Think about how you can leverage machine learning to analyzed data in your system to improve efficiency in your current processes. Be on the lookout for new machine learning applications emerging in SAP S/4 HANA Enterprise Management.

Perhaps a way for you to start on the topic of machine learning is to consider how machine learning can help with tasks that are so tedious that users often avoid doing them. Then think about how machine learning can have a transformative effect on your supply chain execution.

Assigned Tags

      You must be Logged on to comment or reply to a post.
      Author's profile photo Hanine Ouskou
      Hanine Ouskou


      FYI, i'm working on a PoC to use Machine learning to check the inconsistence of the master data. For some fields (correspond to a logic business rule), ML will return value 0 for incorrect data or 1 for good data...

      Author's profile photo Kirupakaran Kannan
      Kirupakaran Kannan

      Hi  Hanine Ouskou .,

      Just curious to know how your POC went with respect to using machine learning for master data inconsistencies. where you able to achieve the requirements ?