Waitrose, PCMS and DMU in computer science link-up

UK grocer Waitrose and one of its software and IT services providers, PCMS, are working with De Montfort University (DMU) in Leicester to solve some every-day retail challenges.

Over a six-week period, 17 final-year and postgraduate students in computer science and intelligent systems explored ways to improve in-store efficiency. The initiative has given the retailer and vendor ideas to take back to their respective organisations, while allowing students to apply their skills to real business cases. 

All parties are keen to continue the partnership, with Stuart Eames, retail innovation lead at Waitrose, saying he is confident that repeating the process again “would be equally as beneficial”.

Students applied their knowledge of machine learning to four issues commonly faced by food retailers: managing shelf stock efficiently; avoiding incorrect product selection at the tills and during self-service; flagging customer errors while using the ‘Quick Check’ service; and to create better prices for products reaching the end of their shelf life.

Participants worked closely with representatives from the supermarket chain and PCMS, and visited a Waitrose store in nearby Oadby to gain first-hand insight of the problems, before presenting their findings to senior managers.

One student’s ideas revolved around the application of data mining and using reinforced learning alongside different modelling techniques to build a dynamic way of assigning new prices on reduced items. Another used linear regression – a statistical approach for prediction tracking – to help with the reduction algorithm. 

The latter was a concept from 21-year-old, Tev Allen, who calculated the quantity of stock, added in its current shelf life, and dividing that by the number of items sold per hour/week/month. 

"I then applied the data I extracted from the reduction algorithm to the current outstanding issues, providing the rest of the solutions,” Allen explained.

“The experience was a one-off opportunity. It would especially benefit students who didn’t do a placement, as it gives you the chance to work with real-world data and present your work in a professional way to a senior management team.”

Eames added: “Providing real business problem statements to the students was fairly easy, but what was powerful was seeing how important it was for them to have real-life examples to solve and how enthusiastically they approached this.”

PCMS said the presentations had given the company some avenues for research and development which may aid Waitrose and future projects. 

Dr Archie Khuman, DMU’s academic lead on the project, commented: “As this was the first incarnation of the machine learning project, what they presented will undoubtedly provide the foundation for future expansion.

“Working with Waitrose and PCMS has been an incredibly joyous and fruitful experience, and one I will look forward to evolving.”

What’s Hot on Essential Retail?