Bartel van de Walle

Science for the benefit of people. All people. Worldwide.

Story nr.


Previous story Next story

The impact of data on disaster relief

Professor Bartel van de Walle holds the chair of Policy Analysis at the Faculty of Technology, Policy and Management. In his research he looks at the role of data and information in decision-making processes during disaster relief operations. Van de Walle pleads for a more context-aware approach to information sharing: “Without context, without a feeling for what is really going on in disaster-hit areas, things can go very wrong.”

A man walks into in a hotel lobby and shows a contact a photocopied document containing classified information. The contact excuses himself to visit the bathroom where he photographs the document with his cell phone. Two days later the ‘secret’ information is all over the news. Sounds like a spy movie? It is the kind of thing Professor Bartel van de Walle has witnessed visiting disaster-hit areas all over the world. “Information sharing in complex disaster situations is difficult and sensitive”, he says. Complex actually means conflict here, to differentiate from natural disasters. Though natural disasters are complex enough in their own right, there is a clear common enemy, e.g. the earthquake, typhoon or flood that is causing the trouble. In complex disasters other issues play a part too, such as strategic, military and security concerns.

Politically sensitive

As a consultant for OCHA, the UN’s Office for the Coordination of Humanitarian Affairs, Van de Walle once compiled a report on data management during the conflict in Syria and whether the way information was being shared was leading to the right outcomes. “At some point, one of the organisations active in the region was undertaking cross-border food deliveries, before this was authorised by the Security Council. Some of these deliveries may have ended up in IS-occupied areas. It is in this kind of politically sensitive setting we are working, where information has a very real impact.”

“Data are sensitive”, is his message. It is a common pitfall for researchers to lose sight of the potential consequences for people in the field. “More and more, data are analysed over here, visualised on some neat map or graph, and sent back to be dealt with locally. We are getting ever better at that analysis, but it creates a distance to the local populations dealing with disaster. It is also the easier part of the work, and can give a false sense of having achieved something, “ Van de Walle argues. Academics should step out of their clean lab environment and go and experience what is going on on the ground.

Street credibility

This is exactly what he and his fellow researchers are doing in this still small, but growing domain of expertise. Among others, Van de Walle has been to the Philippines, Syria, Nepal, Haiti, and Ebola-stricken parts of Western Africa, but also to Geneva and New York, where the decision-making processes often take place. Once in a disaster-hit area, the reality is that rescue workers are generally not too keen on taking time out to talk to researchers. “You can’t just fly in and announce you are a researcher. You need street credibility. It has taken us time to get to know the parties involved and to build up a relationship.”  One of the ways they do that is by dividing their time between offering practical assistance on the ground and gathering information. “We try to help people all across the hierarchy, from those sleeping in tents in the mud to the directors leading humanitarian operations from their head offices,” Van de Walle explains.

Often, much can be improved on an organisational level, as was the case in Syria. “Surrounding countries were trying to coordinate relief from their country offices, but it is the New York office that briefs the Security Council. For that they need data from the country offices, who were getting overburdened by all the information requests to the point where they could not do their job anymore.” On Van de Walle’s advice procedures were changed so the information needed was first prioritised and local agencies were only asked to provide the absolutely essential. “That is the kind of policy change that really affects operations.”


More technology is not always the answer, even though providers like to think so. “I was once in a meeting on the Ebola crisis, when Mark Zuckerberg himself phoned, offering to donate 30,000 smart phones. Not such a good idea, considering the lack of connectivity in the area,” Van de Walle muses. Even with successful technology things can still go awry. Take the Ushahidi crowdmapping platform. Ushahide, meaning testimony in Swahili, was developed to map post-election violence in Kenya in 2008. “The initiative was a success and was later used for disasters all over the world, including the Us Deepwater Horizon oil spill and the earthquake in Christchurch in New Zealand.”

Van de Walle’s aims to bridge the gap between technology and humanitarian field work: “We work at the intersection of practice and science. We want to make technology developers aware of the potential consequences, so that what they come up with does indeed benefit aid workers. We also have to ask ourselves if people in the field truly need all this information. Analysing data is an easy way to convince ourselves we have achieved something,” he says. At the same time he has to ensure his work holds academic validity. “It is a bit of a balancing act. Sadly, you still get more kudos for writing papers and developing cool stuff, than for getting your hands dirty doing fieldwork.”

Fast response grants

However, getting research on disaster response financed is still another matter. “In the US the National Science Foundation offers fast response grants. No large amounts of money, but enough to cover travel expenses. Yet, although the EU is the largest donor to aid operations in the world, the research budget is negligible.” There is all the more need, because Europe is currently facing its own humanitarian disaster, the refugee crisis. “The EU has no clear picture of who all these refugees are, where they come from and how many of them are there. We once organised a workshop on refugees and data to work on this, but it was never followed through because of lack of funds. Yet the crisis continues.”

Van de Walle has noticed however, that it has finally dawned on organisations like the UN that more research is needed. He wants to use that momentum to lobby for more funding. Part of that effort lies in convincing aid organisations that this kind of research can benefit them. “Our work often involves pointing out where and when things tend to go wrong. That is not always appreciated by hardworking agencies. We have to make sure they see us part of their team, making a real contribution. I believe every humanitarian response team should have its own researcher.” Until then, some lateral thinking can help: a trip to the Philippines in the aftermath of the 2013 typhoon was financed by crowdfunding.


Nevertheless, a number of projects are on the way or in the pipeline. “We work a lot with the Red Cross, and TU Delft has recently signed a collaborative agreement with the Dutch Coalition on Humanitarian Response.” One such project is iTrack, in which twelve organisations from eight countries are developing a tracking system. Van de Walle: “Cases of aid convoys being attacked, or aid workers kidnapped, are on the increase. We are working on system that can track such convoys, but that can also predict what would be the safest route to take.” Here too, there are potential risks to consider. “Such a tracker would contain a lot of information, so what happens if these fall into the wrong hands?” Van de Walle remembers a journalist in Sri Lanka who had an Ericsson phone that had a single, large reset button. “If you got caught with the wrong numbers on your phone, things could turn nasty.” Perhaps in the wake of the relaunch of the Nokia 3310, this Ericsson feature can also make a comeback. After all, it is not about high-tech, it is impact that matters most here.

Story nr.


Previous story Next story