Mass migration is a modern-day reality. Over the past decade, the global population of forcibly displaced people has grown to a record high, from 43.3 million in 2009 to 70.8 million in 2018. In real terms, this equates to one person becoming displaced every two seconds. Whilst many of these displaced people move within their home country - 25.4 million people in 2017 — become refugees. 

The response to a refugee crisis is often reactionary. Temporary camps are erected once thousands of people begin to move and aid agencies are flown in. But what if it was possible to predict these migrations before they reached crisis point? Would the planning this allowed provide the ability to save lives or facilitate a less traumatic experience for the displaced community?

Rana Novack, an advocate for refugees has helped develop ‘IBM’s Refugee & Migration Predictive Analytics Solution’ a first-of-a-kind programme that leverages machine learning and cognitive computing to enable Government agencies and humanitarian aid organisations to better manage refugee and migration crises. 

Novack, the daughter of Syrian migrants, set out to counter what she believes is the consistent failure of the international response to refugees in crisis. Whilst helping family members gain passage to safety, like the 13 million Syrians that have left since 2015, she found that while the agencies knew that a massive crisis was unfolding in Syria, there was no plan. Families had to make a potentially illegal, expensive, life-threatening border crossing before they could get any help. 

Through her business development role at IBM, she understood the potential of modern computing’s ability to predict future scenarios, and set out to apply it to the refugee crisis. The software is currently being tested in prototype form, but if successful, it could provide a potentially life-saving advantage to policy makers, humanitarian groups and national governments – ‘the ability to plan ahead for a refugee crisis — before it happens’.

“The idea of being able to predict a crisis, to proactively respond — it’s not rocket science…If they can tell me what kind of shirt I want to buy, there must be a way for us to predict the refugee crisis” She explains.

To build the software, Novack gathered other IBM employees to understand how far machine learning could take the idea of predicting refugee crises. One of those was IBM researcher Rahul Nair, who was studying how people move within cities in order to design smarter bus networks. He then applied the knowledge to predicting displacement and found that when using historical data, he was able to predict the last major migrant crisis’s to within close margins accurate enough to help make decisions. 

Novack explains that: “If we know the pathways that people are taking ahead of time, then we could reinforce those pathways and save lives. Why do we wait years into a crisis when thousands of people have drowned, hundreds of thousands are killed, millions are displaced, and only then we talk about opening our doors?”

The team found that although there are many factors that influence whether an individual will decide to leave their country and where they’ll go, analysing these “push-pull factors” can allow a fairly reliable prediction. To identify these factors, the team studied international news on migration since 2010 alongside development and economic data from World Bank since the 1960s and the last 15 years of migration numbers as sourced from the UNHCR, mapped against logistical factors such as the distance between countries, and cultural factors such as language and colonial links. This data was fed into data models that used supervised machine learning to identify the mathematical relationship between these causational factors and the resulting effects. In doing so, they were able to identify “Push factors” – circumstances that pressure people to leave their home countries, such as unemployment, conflict and violence and “Pull factors” - ideas that attract migrants to a particular nation, including perceived economic opportunity or an existing immigrant community. 

To ensure the software was able to learn and adapt its predictions as fast as real-life situation can change, the team included the ability to add scenarios like a policy change or border closure to see how migration numbers are affected – such as when Hungary closed its border with Croatia in October 2015. Once it understood the relationship between factors, the programme was able to calculate migration numbers for the year ahead in a similar way that an algorithm is able to produce a long-range sales forecast.

The model also had tools added that would enable short-term forecasting. This allows agencies to plan for refugee movements based on a short-term weather forecast, for example, allowing organisations to make decisions about moving resources and staff among the refugee camps they serve.

Ultimately, this data-based model provides insights into migration that aren’t captured on the ground. By looking at the larger picture, the IBM team can help support aid and help direct it to where it will be most needed. The approach has been welcomed by the NGO community and several collaboration projects are in action to either help serve the refugee community or collect data to improve the model. 

For example, the IBM team is collaborating with the Danish Refugee Council (DRC), to survey Ethiopian refugees on their way out of the country about why they leave to allow the model to better predict the movement of these refugees by understanding which pressures are more important than others, and how exactly these forces interact to make a person leave. 

Novack’s aim is to predict the next refugee crisis before it happens so that better strategies of help, support, providing governments and aid agencies with time to set up safe ways for people to leave their countries before they’re desperate enough to turn to smugglers and human traffickers. 

The IBM solution demonstrated the potential that tech has in helping solve some of the world’s greatest challenges, yet humanitarian organisations often cannot afford to develop such sophisticated software. Explaining the issue, Novack said: “There’s an innovation gap in the humanitarian sector. We have to make sure that the people who want to do the right thing have the tools and the information they need to succeed.” It is therefore vital that innovators from anywhere in the world are empowered and supported to deliver their ideas so that we can help those most in need and nurture global prosperity. 

Contact Us

If you have any questions