Joel Kaiser is a project manager and humanitarian professional with Medair, a humanitarian organization inspired by Christian faith that works to relieve human suffering in some of the world’s most remote and devastated places. The group provides a range of emergency relief and recovery services, including health care and nutrition; safe water; sanitation; and hygiene, shelter and infrastructure. Kaiser has recently been exploring the use of unmanned aircraft for humanitarian relief.
Could you give us some background on Medair?
Medair is a Swiss-based humanitarian organization. We differ from NGOs [nongovernmental organizations] you might be familiar with in that we focus exclusively on humanitarian crises, both natural disaster and man-made crises. We specialize in three main areas of humanitarian assistance: shelter and infrastructure; health and nutrition; and WASH, which stands for water, sanitation and hygiene. These are the three areas of expertise. And what we found over the last several years was just how much the use of information was critical to delivering excellence in those three core areas, particularly getting your hands on accurate data and then making better informed decisions based on the data.
So before you had to go in blind and hope for the best?
No, but there has definitely been a slow revolution taking place where you can get your hands on a lot more data than you could even five years ago. … A good example is Nepal [after the 2015 earthquake]. We got to Nepal, and there are these big numbers, like in Sindhupalchok District we’ve got 95 percent destroyed or damaged buildings. That’s an overwhelming number for planning and decision making. Where should we go first in Sindhupalchok? Once you actually get there, it’s just too big of a number to work with. We need to know more precise data, down to the village level. … How do I collect that data quickly, efficiently in order to make a quick, efficient decision of where to prioritize the relief?
Is that how you started thinking about UAS?
Yes and no. That dilemma is part of what led us to explore ways to collect and analyze data better. That topic led us to UAVs. Indeed, we first started thinking about UAVs due to the frustration of trying to use GIS [geographical information systems] for spatial analysis on, say, a refugee camp but not having an up-to-date base map for the camp.
So you know, for example, a place like Lebanon, where we work with Syrian refugees in the Bekaa valley, we know there’s hundreds of thousands of refugees that are there and coming across the border through the course of the year. So in order to anticipate a need in these hundreds of little tiny refugee camps, we started to collect data on how people are arriving, when they’re arriving, and looking for patterns in that data and then overlaying that using GIS.
Our frustration was that at the base layer was outdated. It was just an empty farmer’s field, because we had no current imagery of the field with refugees camped in it. So we lost the most important aspect of that process, which is the spatial analysis — looking at that camp, the width, the breadth of the camp versus population density. And there’s lots of health vectors you can run off with that, so it was a frustrating thing. It was like, if only we had an up-to-date aerial image of that camp. Hence, our exploration of UAV technology.
By the time you started using them, had the perception that they weren’t just military systems started to change, based on the size of the vehicles?
I think so. Certainly there are still instances with the government where we’ll mention UAV or drones, and they’re like, “Whoa, what are you talking about?” And then once we show them the platform we’re actually working with, things change. For example, in the Philippines we were sort of in a limbo, because we had written all of the official letters to the military and the government, the civil aviation committee, and others asking for permission to fly. But they were not making a decision. Finally I literally went to their offices with the SenseFly eBee, and then once they saw it they were like “Oh, that’s the drone that you’re talking about? It looks like a toy! Go ahead and fly.”
What are some of the hurdles that you and other NGOs face when it comes to using unmanned aircraft?
What we’ve found … is the initial investment for the type of UAV that would really help us is too big of a bite to take when it’s going to be applied to a project or coming out of a project budget. By that I mean, aid budgets are always tight. There is no room for big investments. And if it’s not coming out of a project budget, it’s going to come out of the NGO’s standing capital, and that is competitive to use among the various departments.
There are a lot of different technologies it can be applied towards. The thing with drones right now, and I can admit that, it’s still sort of in its experimental phase. As of right now, I cannot draw a direct line from the use of UAVs to the better delivery of aid into the hands of beneficiaries. But I anticipate this will change soon as prices come down and more use cases are identified. Whereas there’s other technologies, particularly mobile digital technologies, where you can go into a village with a tablet and in real-time crunch the numbers you’re collecting, … this is a more direct line to better outcomes in aid.
What would the perfect drone look like, and how much would it cost?
What you generally find in a disaster response is a bunch of relief agencies struggling to prioritize affected communities for the delivery of aid into them. A data set such as that would help me identify whether I should prioritize village A over village B. This type of data, on the macro level, keeps getting refined and improved as more information becomes available and more time passes. In fact, if you analyzed all the past recent disasters, you could draw a curve of information versus time. So, its our hope to find ways to beat that curve and get to the most affected villages faster.
And how do we do that? Well one way is that we try and make that process as efficient as possible, such as mobilizing local volunteers and sending them out with mobile phones to remote communities to have them report back in real time. Obviously, there are some communities that are too far away, or they are unreachable due to landslides, etc. I’m cautiously optimistic that drones can play a major role, but to do that, the drone
needs to be able to go where we can’t, and therefore it’s got to be able to fly long distances and send data back in real time. Plus, the safety has got to be top-notch, since one mistake and it would be a showstopper. And, here’s the clincher, this system has got to be affordable to a humanitarian organization. And what is that magic number? I don’t know yet. But I know a SenseFly eBee is probably about the top end right now. But an eBee doesn’t get me beyond line of sight; it doesn’t get me real-time data feed. So my plea is to a donor or a manufacturer out there that can solve this problem with us.
So you’ve been working with drones for three years?
Yeah, we started in [the] Philippines after Yolanda [known internationally as Typhoon Haiyan] that was in September 2013. That was our first deployment. We’d been thinking about it six months’ prior because of the situation in Lebanon. And so then we started to explore UAVs. And the good thing is our headquarters are based in Lausanne, Switzerland, where you’ve got SenseFly, Pix4D, DroneApps, EPFL [École Polytechnique Fédérale de Lausanne] and a lot of interesting technology coming out of that space, and so we’re able to make those linkages very quickly.
You’re a member of UAViators, a humanitarian network that shares information about UAV operations. How important is it for groups to talk like that?
Coordination, in humanitarian aid, is the lifeblood. It’s mission critical. However, it’s voluntary, … whereas in emergency management, you’ve got an incident commander and a command and control structure, right? The incident commander can bar your access, but you will never find that in humanitarian aid. Humanitarian aid uses a cluster system, if you’re familiar with it, and this is based upon voluntary coordination. Consequently, it’s really incumbent upon those drone pilots, whomever they might be, coming into a disaster response to plug themselves in.
UAViators is a platform which enables drone operators to come together and coordinate their actions before, during and after a response. It’s taken a lot of work to get there, but it’s starting to pay its dividends. For example, the code of conduct, I think, is a vital document. There are also guidance documents on conflict, data ethics, principle partnerships, engaging populations, and I think all of these speak to the culture of coordination in humanitarian aid.