Internship: ‘Robust optimization’ of World Food Programme relief operations

“This was definitely one of the coolest ways to use math for optimization,” Riley Badenbroek says, referring to his internship with the World Food Programme in Rome. He worked there on an optimization model for food aid in crisis regions. Former fellow student Koen Peters, whom he knew from the Outreaching Honors Program for talented students at Tilburg University, had brought him into contact with the WFP. Riley seized the opportunity to gain practical experience as part of his Master’s program of Operations Research & Management Science (currently called Business Analytics & Operations Research).

Enriching experience

“The internship was also an enriching experience for me personally,” Riley says. “I discovered that I prefer solving mathematical puzzles to addressing the business operations side of the problem. When I have completed my work at the WFP, I am going to do PhD research here at Tilburg University, on convex optimization problems, a much more theoretical subject.

My role at the WFP was to expand the mathematical model for food assistance to include uncertainty factors, in other words, to make the optimization more robust. There are many uncertainties in developing countries and in war zones. For example, the purchase prices fluctuate or there is no room in a port for relief ships to dock. The model takes this into account by making a trade-off between risk and expected costs. If there is a large risk, you need to look for safer but more expensive options. I was able to visualize that trade-off in a chart.

Valuable

The best moment was when we presented the model we had developed to the management. It proved to be exactly what they had in mind and they found it very valuable. A very rewarding moment.

I am currently working on the model as a consultant, to make it more user-friendly for the World Food Programme staff via a website. I had website-building experience, so I can make a useful contribution.”