Skip to main content

Efficient Consumer Response (second project)

ECR

Professor Aris Syntetos has been awarded research funding for the project ‘Using machine learning to control inventory record inaccuracies’.

The project addresses the discrepancy between what we think we have in stock (based on our computerised system) and what we physically have in stock. This equates to a significant issue when multiplied across thousands of products typically stored in retail.

Funded by Efficient Consumer Response (ECR), the project will be jointly delivered by Cardiff University, EM-Lyon (Professor Yacine Rekik) and TU Darmstadt (Professor Christoph Glock). It will explore how machine learning can help make more robust inventory control decisions in the context of inaccurate stock records. The project will be delivered with the participation of nine of Europe's largest retailers.

In a previous ECR project, Aris, Yacine and Christoph found that up to 60% of inventory records are incorrect, and that when the inventory records are corrected sales can increase by 4 to 8%. You can read more about the project on the ECR website, and the webpages of the PARC Institute.