“The model helped the WFP save EUR 27 million in 2017”

How big data could help resolve the food crisis

Hein Fleuren of Tilburg University uses data and mathematical models to improve supply chains for aid organizations and businesses alike. The UN World Food Programme already provides aid more efficiently thanks to his work.

This article was previously posted on Rabobank.com on June 12, 2018.

Hein Fleuren, Professor of Operations Research at Tilburg University, has optimized the supply chains of various large companies. His corporate customers include parcel delivery service TNT Express: the multinational saved EUR 200 million (USD 235 million) and reduced carbon dioxide emissions by a quarter of a million tons.

Other corporates hoping to benefit from Fleuren’s expertise will need to be patient: he is currently applying his mathematical models to help aid organizations, such as the UN’s World Food Programme (WFP).

Convincing the WFP

When he approached the Rome-based organization five years ago, it was not immediately convinced that it could benefit from his work. “The WFP wondered why a mathematician had come to see them,” says Fleuren. “It was quite an ordeal to convince it to invest in more efficient logistics.

“That has to do with the way aid organizations are funded – they are expected to spend their money on direct aid, which makes them reluctant to invest in innovations where the outcome is uncertain. And in all fairness, the projects I work on could fail.”

“The model helped the WFP save EUR 27 million in 2017”

- Hein Fleuren, Tilburg University

How to feed more mouths

Despite the WFP’s initial reluctance, Fleuren was so convinced of his models’ potential that he refused to let the matter go. “The WFP is one of the world’s largest logistics organizations, transporting four million tons of food a year. I could see that it had to be possible to make the flows more efficient, so that it could feed more mouths on the same budget.”

With support from Tilburg University, he was able to spend one day a week looking into the WFP’s food aid supply chain. He also spent time preparing the WFP managers for the changes to come by developing a game for them that used food aid flows in South Sudan as an example.

Sorghum instead of wheat

After Fleuren had worked for the WFP for free for two years, it came around. It funded a small program that allowed the professor to work with three students. Fleuren and one of the students, Koen Peters, drew up a complex model that made calculations using data about food, such as price, supply, demand, as well as harvest times, routes and transport options.

“We achieved a breakthrough by changing a basic principle,” explains Fleuren. “Instead of looking at the type of food – wheat, rice or vegetable oil – we started with the nutrients required per person per day. Then we looked at what the ‘food baskets’ should contain to meet these basic needs. That gave us an understanding of more efficient alternatives, like substituting sorghum for wheat. This African grain is cultivated close to emergency areas, which significantly shortens the supply chain.”

The model proved to be successful, and WFP applied it to various food programs in Syria, Iraq, Yemen and Ethiopia. It was often able to feed 10 to 15% more mouths without spending more. In fact, it has cut costs. “The model helped the WFP save EUR 27 million in 2017 alone,” he says.

“The 200 aid organizations in South Sudan each work in isolation”

- Hein Fleuren, Tilburg University

Fleuren still has one item on his wish list: achieving better cooperation between organizations that distribute food. “There are almost two hundred aid organizations in South Sudan alone, each working in isolation. If we could develop a model for them and they joined forces, their combined capacity would really bring significant gains.”

Better food security

Fleuren would also like to develop models to improve food security. “Transporting food is inefficient and creates unnecessary greenhouse gas emissions. A fellow scientist calculated that the average British Christmas dinner travels a distance of almost 100,000 kilometers before being eaten.”

According to Fleuren, models and big data could provide an insight into the opportunities to produce locally for locals. “That would mean reduced emissions and less food being wasted on long journeys, increasing food security.”

“Big data is just like beer”

In fact, Fleuren sees endless possibilities to improve the world using big data. He is involved in initiatives for refugees and the distribution of medicines. Thanks to the big data hype, he finds it easier to get a foot in the door with aid organizations these days. However, its worsening reputation is a worry. “I always tell my students that big data is like beer and electricity: you can do bad stuff with it, but also great things.”