Clients

Rajiv Kumar, <rajiv.kumar01@sap.com>, Product Owner
Abhijnan Saha, <abhijnan2004@gmail.com>, Architect

Student contact

Rohith Mandayam Devasikamani, <rm977@cornell.edu>

Project Summary

The Bills of Material (BOM) topic is one of the core topics of Product Lifecycle Management area at SAP. The BOM consists of a unique header defining the product and a list of items used to create that header. An example is as follows: {Coffee Cup: Plastic lid, green plastic straw, plastic transparent cup, etc.}. In the event that the manufacturer decides to replace one material for another, the algorithm must determine where every instance of the given material is used in the database of BOMs and replace. This functionality already exists.

Our algorithm will advance this framework by taking in the history of customer data and various other features (cost, texture, color, availability, region, etc.) and finally suggesting the best material that could be used to replace the given material by employing Machine Learning. If the preceding functionality is complete, we will add functionality to suggest next best materials on similar lines.

Another additional functionality which we plan to implement is to provide a recommender system for the “Alternative Item”. The “Alternative Item” is a field in the items database table which stores the information of an item that can be used instead (based on availability of the replacement decision). In case the material is not available, the recommender system would suggest an alternative based on the material type.

The project in all, aims to incorporate the principles of Software Engineering while using other tools and technologies involving machine learning, data processing and wrangling, principle component analysis and web technologies.