Drone, meet the humanitarian cluster approach

BANGKOK — The World Food Programme, one of the first humanitarian organizations to explore the use of unmanned aerial vehicles back in 2007, hasn’t officially deployed any UAVs in responses this year. In fact, the agency’s drone count currently stands at zero.

Instead, the United Nations food assistance branch has been putting drones in the sky only as part of a project to establish clearer guidelines for the humanitarian sector at large.

“It’s something internally we agreed. We don’t have a guideline for using this new technology. There are a lot of questions around privacy, data protection, perceptions of the communities, accountability to affected populations … so we decided we are not going to deploy any until we sort this out and establish a practice,” said Haidar Baqir, information technology emergencies officer for the WFP’s Asia-Pacific Regional Bureau.

Several country-level communities of practice for the humanitarian use of drones already exist around the world, with nongovernmental organizations as partners to share the costs of flying drones and the data produced, often following a disaster. In Nepal, Tanzania, and Peru, “robotics for good” company WeRobotics has established “flying labs” to transfer relevant skills and technologies to local partners to incubate new drone businesses.

The Belgian-funded WFP initiative that Baqir is heading up aims to collate those ideas to design solutions for the coordination of UAV use in the broader humanitarian community, addressing safety and data protection risks along the way. The result, Baqir hopes, will be an official partnership framework and accompanying code of conduct by early 2018 whereby drone deployment in emergencies will be managed by one humanitarian aid cluster in a transparent way.

Localizing drone use

UAVs are increasingly being recognized by humanitarian organizations for their potential effectiveness in disaster response, helping teams safely observe situations in real-time during emergencies or quickly collect damage assessment information in remote areas. Drones can also provide “Wi-Fi in the sky” to get communities back online following a disaster, an idea the WFP has been experimenting with.

At the same time, overuse or improper use of drones following disasters in the past few years have left many aid groups wondering how to move forward responsibly.

Several organizations already exist to help educate people about best practices, such as WeRobotics — which is partnering with the WFP on the drone coordination initiative — and the Humanitarian UAV Network, both founded by humanitarian technology and innovation expert Patrick Meier. But the WFP project is the biggest effort on the international stage to professionalize drone coordination throughout the sector — and it can’t come soon enough, according to Meier.

The response to Nepal’s 2015 earthquakes, for example, was a mess “from the drone side because you had a bunch of drones self-deploying and causing much more of a headache than anything else,” Meier said. “The pilots, many of whom flew in from Silicon Valley, had never worked in Nepal before, let alone in a developing country … it was a disaster.”

Aside from the lack of transparency and general disregard for government permission, the drones were collecting data that were then likely loading onto a server that only one organization would access, creating a duplication of efforts the humanitarian sector would like to see eliminated from emergency response. After witnessing this firsthand, Meier said, he went to the U.N. Office for the Coordination of Humanitarian Affairs and the WFP to warn them that this “is only going to get worse unless we start building local capacity and developing the coordination mechanisms that are necessary to do this safely and responsibly.”

It’s one of the reasons Meier established flying labs in three strategic countries, he said: “To provide capacity locally so you don’t have to hire an American, European, or Australian drone company to fly halfway across the world to do some mapping in Tanzania, for example.” Two more labs are scheduled to open, in Panama and Senegal, next year.

In the Philippines, drone company SkyEye has been providing aerial mapping and disaster response surveying services for several years. Originally, “NGOs would ask us to fly in, then we’d fly back out, then another NGO would say, ‘Can you fly back in?’” said founder Matthew Cua.

Now, they’ve formed a community of practice, a loosely based organization of 74 local and international NGOs in the Philippines who share costs and data. “Instead of having their own departments, we’re a department externally shared among everybody,” Cua explained.

Following a disaster, SkyEye might map affected schools for Save the Children. World Vision might overlay data on damage per house while the IOM surveys those who may have lost homes and will need to move into alternate housing.

“With one data set we are able to pull enough metadata out of it that we can share it among ourselves to be more efficient,” Cua said.

A few big hurdles

Avoiding overlap from mass drone deployment — and another Nepal earthquake drone experience — is what Baqir hopes to achieve with the WFP initiative, as well as tapping into lessons learned from smaller communities of practice like Cua’s and, perhaps most importantly, getting wary governments on board with the potential positives of UAVs for disaster response.

The single biggest challenge to drone use for humanitarian aid remains local regulations that were usually created for commercial drone use, according to Meier, who saw this again recently in the South Pacific while piloting drones for the World Bank’s UAVs for Resilience program.

“The regulations are not made with humanitarians in mind,” he said. “We have time pressures that maybe a Hollywood film team, for example, won’t have. So it just slows everything down.”

It took the WeRobotics team more than six months to get permission to work in Nepal — even though it was partnering with a local government ministry, said Meier, who is calling for greater interaction and advocacy between senior humanitarian aid leadership and local civil aviation authorities.

In a few weeks, the WFP’s Baqir is headed to Mozambique to conduct the last of four workshops and drone simulations in disaster-prone countries conducted as part of the initiative, which has already staged workshops in Myanmar, the Dominican Republic, and Peru. The goal of the workshops is to engage with local government bodies — especially natural disaster management authorities — and help them create a roadmap for enabling safe humanitarian drone deployment.  

“We ensure that all the actors in country are coming to these workshops to look at what the challenges are, including local NGOs, government, universities, private sector, and sometimes donors as well,” Baqir said.

Another challenge of drone use is identifying how to integrate drone technology with overall satellite imagery assessment, which takes place immediately following disasters. Drones are particularly useful when it’s cloudy, for example after a hurricane, when satellite imagery in certain areas is impacted by weather. But without standard data sets — such as flight height, resolution, and camera angle — it won’t be possible for drone data to complement existing data collection efforts.

A new model to define future humanitarian drone use

Once standards and regulations have been set, it’s time to look closer at coordination, which is the biggest mandate of the WFP’s current project.

Ideally, Baqir said, drones would live under the Emergency Telecommunications Cluster, which is led globally by the WFP and is one of 11 clusters that identify key humanitarian sectors during emergencies. Requests for information would filter through the cluster, which would already have established partners it could call on for in-country coordinators, drone pilots and data analysts as needed, and be responsible for the government permissions process.

If a request is deemed suitable for drone technology use, the cluster would then coordinate flights to avoid overlap.

Though the project is still in talks over what to recommend for data sharing, “We need to have a flexible system that can communicate with other platforms that already exist,” Baqir said. One idea is that data would be published on, or securely linked to, the Humanitarian Data Exchange platform.

In December, Baqir hopes to present a model for a partnership framework that could later take the form of a charter, where all humanitarian actors would agree to the cluster drone approach and to the minimum requirements for partners to use UAV tech in the sector.

The Humanitarian UAV Code of Conduct, established by Meier’s Humanitarian UAV Network, is something Baqir sees organizations adapting and becoming signatories to after a period of feedback. The project also hopes to present global humanitarian UAV regulation that countries can choose to adopt in their contexts to more successfully enable emergency usage of the technology.

“We hope by next year we are deploying drones not for the sake of deployment, but because there is a need and a gap … And that we have an established practice rather than an ad hoc situation; that is our goal,” Baqir said.

Read more Devex coverage on technology and development.