// Camp Forecast improves supply management for refugee camps

Each year, 1.3 billion dollars in humanitarian aid is lost in refugee camps due to poor supply management. As a result, an estimated 1.88 million children, women and men each year do not receive the essentials to keep them safe and in good health. In collaboration with Elva, we developed the Camp Forecast tool for IOM to improve forecasting requirements and thus save time, reduce costs and possibly even save lives.

Fluctuating population in refugee camps
IOM has decades of industry-leading humanitarian experience and ELVA has experience collecting data in 20 conflict-affected countries around the world. Together with IOM, we focused on a crucial problem that affects many refugees worldwide: the challenging supply management in refugee camps. Many refugee camps have poor IT and road infrastructure, skilled personnel are scarce and the local government and population are not always helpful. In addition, the population of refugee camps fluctuates, which creates great uncertainty about the demand for supplies. This makes it very difficult to predict how many clothes, blankets, medical supplies, tents, etc. will be needed in a given month.

Partly as a result of this, there are many shortages, as a result of which many refugees miss certain types of basic humanitarian necessities. It can also happen that there is too much stock, which puts a strain on an already overloaded supply chain and increases storage costs. Better forecasts are needed to reduce both shortages and oversupply.

Data on IDP-camp population
Using historical data on IDP camp population and Machine Learning, supply and demand can be predicted more accurately. Better forecasting and scheduling reduces inventory shortages, overstocking and shelf life, which in turn saves time, lowers costs and can even save lives. In addition, it facilitates accountability to, for example, donors. 

Camp Forecast tool
We have developed an online tool that uses Machine Learning to generate monthly delivery forecasts for IDP camps. We use the following data for this:  

1. IDP population size. We use existing data on the population size of each camp, provided by the IOM. The data is available online in the form of so-called Site Assessments. Our Machine Learning algorithm collects all historical IDP camp population data from these files and then predicts the population size per month based on historical growth trends and seasonal fluctuations.
2. We then multiply the expected number of people within a camp in a given month by the required number of supplies per person. The required number of supplies per person are based on internationally agreed standards, the so-called SPHERE standards. We now have the total number of supplies needed in a camp.
3. Then we check which supplies are already present in the camp, for example communal toilets. This data is entered manually by the camp manager when using the tool for the first time. We'll subtract these so we don't order supplies twice.
4. Finally, we multiply the required stock by the price per delivery. This data is entered manually once and then saved. 

This then gives us the total predicted amount of supplies needed and the total cost.

Faster and more accurate prediction
We have built a prototype, which is a first step towards that solution. This tool helps IDP Camp Managers to forecast the demand for humanitarian supplies in their camp in seconds, without much effort.

It calculates the supply needs and the predicted population size. It then automatically shows which supplies are needed for this population and what the total cost of these supplies will be. We pride ourselves on the fact that the tool can more accurately predict supply needs in less time than it would take to write a basic forecast on the back of a napkin. 

Long term vision
In practice, our vision is that in 10 years' time inventory management in most IDP camps will be carried out in the following way: 

27,8 billion dollars in humanitarian aid
This approach can yield huge benefits. In 2017, humanitarian aid amounted to USD 27.8 billion. If we could increase efficiency by just 0.1%, that would result in more than 27.8 million in funds that could be put to better use.

With our tool, we have a modest starting goal: to reduce deficits by at least 5% in all camps where it will be used. This means that thousands of people within these camps would no longer lack basic facilities. 

Future

1. Gather feedback from IOM, other humanitarian organizations and beta testers
2. Using the tool in at least two IDP camps
3. Scaling up to other countries
4. Using satellite imagery to predict the population of refugee camps.