Global Academy of Agriculture and Food Systems

Roses are red violets are blue, can industrial maths solve the problem for you?

Jess Enright, Lecturer in Computational Methods & Agrifood Systems

At the end of April, I travelled to Nottingham for a Clean and Sustainable Growth Study Group (


I’ve been to several of these sorts of events, and really enjoyed them.  At a mathematical Study Group, representatives from industry bring problems from their businesses that could be addressed using mathematical sciences, and mathematicians from PhD students to full Professors work together on them for several days.  If you’re organised and clever, then you document your work as you go along, and after the event each team produces a short report for sharing with their industrial partner.


Industrial partners provided outlines of their proposed problems ahead of the event, and one immediately caught my eye:


Luckily, this problem also attracted several other enthusiastic researchers, and Team Flowers was formed!


We were very lucky to be working with a rich and interesting dataset: MM Flowers had brought us data on their quality control across all suppliers, and detailed extra data on one particular farm, including the farm’s quality control process and rainfall data.  Team Flowers worked furiously in at least three different programming languages to extract information from all of these datasets, finding associations and patterns from the farm through to the retailer. Many of our findings supported intuition that our industrial partner had been building over time: identifying some flower varieties that are more susceptible to damage to others, showing a relationship between extreme rainfall events on-farm and subsequent fungal damage, etc.  

Jess and colleagues holding flowers


Study groups can be tiring and intense events, but on the final day we were still smiling.  



While the the event is over and the final presentation has been given, our work is not over.  Over the next month or so we’ll pull together a report in collaboration with the industrial partner, and hopefully explore a few promising next steps to support their quality control process.