And a good thing, too. On the planned date of 5 June, there was a big storm, which would have thwarted a successful operation. Thanks to Sweeney, 6 June rather than 5 June has made it into the history books as D-Day.

Using data for humanitarian aid

Using data to change the world in a positive way: that’s the mission of Analytics for a Better World. At the last edition of their annual conference, guest speakers from organisations such as Save the Children, Doctors Without Borders and UNICEF talked about the important role of data in humanitarian aid. Robert Monné, Managing Director of Analytics for a Better World, explains as follows: ‘Right now, data is often used by companies to double their growth figures. The job of a data analyst tends to focus on that aspect, whereas I think it should be about how we can use data to make the world a better place.’

Acting preventively

There were plenty of inspiring examples during the recent conference held on 14 May at Startup Village Amsterdam. Ralf van Otterdijk from Doctors Without Borders, for instance, demonstrated how a smart combination of data can help predict the location of a malaria outbreak. Climate change often creates new sources of malaria. The model that has been developed makes it possible to take preventive action by prescribing anti-malaria medication for people in high-risk areas. Previously, such action was only taken when many people had already fallen ill. 

Child Atlas

At Save the Children is another organisation where data is an important means of extending humanitarian aid. Aranka Hetyey, Head of Digital Enablers, talked about her contribution to a tool that makes it easier to run campaigns. She says: ‘There are umpteen ways to measure the effect of a campaign, but not everyone is good with data. That’s why we went looking for examples that anybody can use, even people who aren’t digitally literate or who don’t always have access to the internet. We explored the elements that would help us tell our story as effectively as possible and incorporated them into templates that can be used by everyone at Save the Children.

‘And then there’s the Child Atlas, a comprehensive map indicating where the most vulnerable children live. The atlas has been put together using a machine learning model, with satellite images playing a key role in its construction, for instance, by showing if there are any roads and the condition they’re in, how far it is to any medical facilities and so on.’

Deforestation

The significance of satellite images was also addressed in discussion panels. Liv Toonen, Technical Project Lead at Space4Good, relies on satellite images to map areas of deforestation. She emphasises the importance of monitoring with data before things go wrong: ‘We have become increasingly proactive in the way we work. If we think the images show unusual activity, we immediately send a message to those in charge of enforcement.’

Context and data equally important

Dick den Hertog, Professor of Operations Research at the UvA Amsterdam Business School, stressed the importance of combining discussion with data analysis. One of the things he did was to identify how many people in East Timor live within 5 kilometres of a health care centre and he used data to draw up a list of places in Vietnam where health care centres should be set up for people who’ve suffered a stroke. Den Hertog observes: ‘Sometimes we have to use data that’s obsolete and sometimes we have only limited access. So it’s always essential to get information about context. That’s a scientific challenge and responsibility. There’s a lot we can predict and yet there’s also a lot we’re unable to do at this point. But we’re getting better and better at linking data and using this to predict such things as conflict.’

Digital humanitarianism

The final keynote was given by Prof. Marc van den Homberg, Scientific Lead at 510, an initiative of the Dutch Red Cross. He shows how the use of analytics can lead to better predictions of the places most affected by a natural disaster. The combination of various kinds of data can improve the predictive accuracy of the places and the people most impacted by a disaster so that aid can be provided more quickly. Together with his colleague Kyriaki Kalimeri, he developed a dashboard for decision-makers to predict the damage caused by a cyclone. The model they used can also be used for other decisions.

At the same time, Van den Homberg warns against the unquestioning use of a model, especially when there’s a reliance on AI. Van den Homberg explains: ‘There are always some assumptions, even if you work with AI. You need to take them into account in your findings and remain critical when presenting outcomes.’ Kalimeri adds: ‘You need to be very careful with figures even if they tell you what you’d like to hear. I once had a student who came to me with a map displaying the biggest humanitarian catastrophes. North America and Europe were right at the top. But this turned out to be the result of incorrect data.’

Work to be done

The Analytics for a Better World conference made clear that there’s still a lot to be gained from the use of data, by working closely together and motivating scientists to deploy data for humanitarian aid. Managing Director Robert Monné can’t wait to expand his mission. ‘We’ve now trained 200 people, but the sector will need 3.5 million skilled workers in the coming years. The potential and demand are huge, requiring scalable solutions. We invite researchers and companies to join our mission.'