Filed under: Development 2.0 Disaster response Social innovation

Here at UNDP we’re in the business of this all-encompassing yet often quite vague idea of ‘resilience’.

For me what it boils down to is this:

How well are the state and its citizens prepared to react to emergencies? And what can be done to help reduce the risks to human lives and their communities when disaster does strike?

Winter is coming

In the former Yugoslav Republic of Macedonia we’ve had some pretty harsh winters as of late: heavy snowfalls resulting in some of the more remote mountain villages getting completely cut off. Things have gotten so bad that helicopters have had to be dispatched to provide emergency relief and assistance.

In some cases, the helicopters landed in villages only to find no one there to accept relief. Some villagers had already left the mountains for villages in the valleys below. Others hadn’t lived there in years.

Because here’s the rub: The last national population census was undertaken in 2002.

There is thus a major gap in data about the population. What are current regional populations and how might these numbers change from summer to winter? How do seasonal shifts affect employment, traffic or an emergency?

We need accurate and up-to-date data so we can ensure a feasible analysis and assessment of the population. What that really means is finding out who is most vulnerable to disasters where and when.

Currently, we’re working with the Macedonian Crisis Management Centre to establish an integrated risk and hazard assessment of the country, covering over 80 municipalities.

But that won’t be enough.

This work is hindered by old statistical data and a shortfall of accurate information about the dynamic characteristics of the population.

We need a tool which can capture the changing dynamics of the population: mobility, habits, vulnerability and when and how these patterns can put different areas at higher risks.

Calling for a solution

This country has very well developed mobile networks. Official data show an astounding 99.8 percent coverage of the population in the national territory with a mobile signal.

The total number of active SIM cards exceeds 2,200,000. This is higher than the total population of 2,022,547!

We can use mobile phones as proxies to see:

  • Whether we can track and better understand how people move around the city in order to establish patterns of mobility and determine possible exposure to emergencies;
  • To what extent men and women use mobile phones differently and whether this could give us some clues as to their social network, support structure, and vulnerabilities;
  • How the numbers of citizens in any given region or city ebb and flow depending on the season;

In short, our hunch is that the way people use their phones can tell us a lot about what type of risks they are exposed to and may give us clues on how to reduce them.

Big data and you

There has been a lot of work on using this type of (big) data for a variety of purposes – from figuring out real-time inflation pressures and food prices to understanding public perceptions about women’s role in labour market to misconceptions about vaccines among parents.

But we haven’t come across a team looking to apply these data to disaster risk reduction (and if you’re out there, we’d love to partner with you).

Data collected during a disaster or emergency, meanwhile, will help reveal different mobility patterns. Comparison of these data during an emergency and during a normal day will help us better understand people’s behaviour and potentially predict their situation and possible needs.

To that end, we’ve managed to get the mobile phone operators in the country interested in providing information on how citizens use mobile phones – which, as far as we know, is another first.

Friends and family

This was terra incognita for us and we’ve been inspired by the remarkable work of Patrick Meier and the Qatar Computing Research Institute.

Other sources like a project in the Netherlands which measured rainfall with phone network signals and using big data from social media to map relief efforts during Hurricane Sandy have helped inspire us along the way.

The UN Global Pulse’s Mobile Phone Network Data for Development also served as a valuable primer.

Where do we go from here?

So here’s where we are: We’ve made some early steps. We got the mobile operators on board. We got the legendary Patrick Meier and the Global Pulse team too.

The future is now and we can’t wait to see where this goes.

We’d love to know your thoughts. If you’ve done something similar, or have any ideas on how we could optimize this idea – we’d love to hear from you.