Using big data in refugee crisis
WHEN SPEED IS KEY: AN APP PEOPLE CAN UPLOAD INFORMATION WITH AND SEND IT TO A RESPONSE TEAM Statistics and analysis may sound dull, but now data is being used to save lives in the world’s humanitarian hellholes.
Once it was a national park. Now a vast area of Bangladesh has been turned into a sprawling refugee camp. Hundreds of thousands of Rohingya people have poured across the border from Myanmar in recent months, forced out of their homes by a brutal army crackdown.
The vicious persecution aims at eradicating a Muslim presence from the primarily Buddhist country of Myanmar. Thousands have been killed for no other reason than that they are part of a community. Aside from the Rohingya, the brunt of the anguish has been borne by Bangladesh, which has welcomed the refugees into a country that can barely cope with its own problems.
Aid workers have poured in from around the world to help, but that has sometimes only added to the confusion.
“How do you deal with an emergency in the chaos of a million people milling around?” asked Leonard Doyle, head of media and communications for the United Nations Migration Agency. “International and local agencies go piling in, installing tube wells next to water points that are contaminated. That’s not smart.
“Our role is to coordinate the response, which is a massive challenge, given that everyone is doing what they want. We have feedback channels and information points to help coordinate such disasters, but when information is collected out in the field where there is no internet connectivity, and only submitted a few days later, it does not have the immediacy or urgency needed.”
The agency finally turned to big data – the science of collecting and analysing large amounts of data, and using it for better decision-making. It developed an online platform to receive the information, as well as a software tool for people to collect and submit information.
“It’s a very simple app that allows people to log information and upload it to a response team, and view it on a mini-dashboard with quick statistics of all the feedback collected. It is easy to synchronise with a community response map, and data can be exported from the platform and shared with agencies via PDF and Excel, live data and infographics,” said Doyle.
“It’s a very simple tool to collect information in the field. Now, information coming from these desperate people gets quickly fed into system. For example, if someone finds a boy who has lost his parents, and inputs that information, it creates a response procedure that ensures the boy is looked after immediately. We need better ways of getting aid to people, and this is one way.” Doyle was speaking at the SAS Analytics Experience 2017 conference in Amsterdam, an event that draws both on 40 years of pioneering data science at the SAS Institute and on some of the most current case studies and strategies for turning data into decisions. Addressing humanitarian crises was a strong theme. “The human being and mathematics are merging,” said Jon
How do you deal with an emergency amid the chaos of a million people milling around?
Leonard Doyle
Head of media and communications for the United Nations Migration Agency The human being and mathematics are merging.
Briggs, the BBC broadcaster who is also the British male voice of the iPhone voice assistant, Siri. Chairing the conference, he issued a powerful warning: “The danger in relying purely on statistics is that it can dehumanise what is often a very human tragedy.”
But he pointed out that the work of the UN Migration Agency showed how data could save lives on global migratory routes. Also known as the International Organisation for Migration (IOM), the body is currently dealing with 65 million displaced persons, 21 million refugees, and 41 million people displaced in their own countries. Almost a third of this latter group is in Africa.
In one month in the DRC alone, 1.5 million people were displaced.
“These folks are the wretched of the earth,” said Doyle. “Already the human traffickers, sex exploiters, human slavers, are there. The vultures are circling. As these people become exploited and enslaved, there is an enormous danger of radicalisation. Yet, much of the suffering is unnecessary.
“These man-made disasters may feel distant on our TV screens, but they have a habit of coming close to us. You have in your hands and brains and pockets many of the tools that could help enormously in dealing with the humanitarian issue.”
The message was reiterated by a member of the Dutch royal family, Pieter-Christiaan Michiel, Prince of Orange-Nassau, Van Vollenhoven, who is also vice-chairperson of the board of the Dutch Red Cross. “I believe big data can make the world better, more humanitarian and smarter,” he declared.
That was the thinking behind the creation by the Dutch Red Cross of a data unit called 510 Global, named for the 510-million square kilometres that make up the surface of the earth. It is described as a “dedicated hybrid team of data scientists and information managers and researchers who apply their skills across humanitarian activities with Big Data”.
The organisation said: “From visualising and communicating information through interactive dashboards, maps and infographics, our team collects, collates and analyses big data, extracting insights and translating them into data-driven decisions, positively impacting humanitarian aid.”
Prince Pieter-Christiaan presented a case study that is still raw in the memories of the Dutch: the devastation of the Netherlands territory of Sint Maarten in the Caribbean by last month’s Hurricane Irma. More than seven out of 10 buildings were damaged or destroyed. The relief operations were a nightmare for aid organisations, the military and government.
The 510 Global team was tasked with both preparing data before the hurricane hit, and assessing the damage afterwards. “We worked with Google, which was able to predict the path of the hurricane, and first responders were able to share information via Google Maps. We knew the hurricane would hit the island, but we wanted to create an accurate picture of where people lived and map the houses on islands.
“We used satellite data to count houses and see where the roads are to reach them. A lot of illegal immigrants were living and working on the island, living in makeshift buildings. We used crowdsourcing to find how many unregistered buildings there were, and that map was used for the rescue operation.
“We used drones for damage assessment, and volunteers used satellite data to map and colour-code the most devastated areas, to focus relief operations. We also used that for the recovery, to see how many roofs were needed for makeshift buildings.”
That still left aid workers scrambling for resources on the ground, but it helped divert them to where they were needed most.
The prince pointed out that the Dutch Red Cross was 150 years old, but was now spearheading the concept of smart aid. “We have a $25 billion budget, but a $35 billion need. There’s a big gap. We are always short of money. Smart aid pinpoints our smart responders, allowing them to be much more effective by seeing where the relief effort is needed most.”
From visualising and communicating information through interactive dashboards, maps and infographics, our team collects, collates and analyses big data, extracting insights and translating them into data-driven decisions, positively impacting humanitarian aid.