Andrew Schroeder: Drones for Social Good
A conversation about using drones for social good and helping us understand the drone industry.
Radiant.Earth is excited to introduce to you Dr. Andrew Schroeder, the first profile in our new Community Voices series. Andrew is recognized world-wide as a leader in geographic information systems (GIS), data science, and applied epidemiology for humanitarian aid and global health. He’s also at the forefront of using drones for humanitarian response. As the Director of Research and Analysis at Direct Relief and the Co-Founder of WeRobotics, Andrew founded the NetHope UAV working group and co-directs the UAViators Humanitarian UAV Network. His educational background includes advanced degrees in social analysis and public policy from New York University and the University of Michigan.
“You’re going to see more of all of this in the future […] much more emphasis on finding solutions to image data scale and analysis for drone projects.”
How would you describe the global drone industry?
Right now, I would say that the commercial drone industry seems to be in the process of retrenchment and reorganization — the achievement by DJI, the Chinese drone manufacturer, of something like global market dominance, is an industry game changer. This disruption in the market, I think, took many companies by surprise and led to massive and somewhat painful changes at firms like 3D Robotics, Parrot, GoPro, and others. For the foreseeable future, it’s going to be very hard for manufacturers of commercial quadcopter drones, which is the vast majority of all drones sold each year, to compete with DJI.
On the other hand, that has meant a renewed emphasis on the more “professional” side of the market, on improvements to fixed-wing drones, on the integration of services and data with hardware and production of complete workflow “platforms” for drones. Parrot, for instance, has done remarkable things with the creation of prosumer-level fixed-wing drones that integrate well with Pix4D for specific use cases in agriculture and inspection. Likewise, Sensefly continues to make extremely high-quality professional fixed-wing products. 3D Robotics is basically back from the dead as a software rather than a hardware platform. DroneDeploy and others have been quite successful at creating fully integrated drone data workflows. Esri, the GIS software leader, has also been successful in regard to integration of drone imagery workflows into a geospatial analysis. You’re going to see more of all of this in the future — better and cheaper hardware, in some cases customization to specific use cases, more emphasis on integrated services and much more emphasis on finding solutions to image data scale and analysis for drone projects.
How are you currently using drones in your work? What inspired you to start using drones for social good?
WeRobotics was essentially set up to use drones for social good — and not just drones of the flying kind but of the swimming and driving kind as well — in order to help organizations, governments, communities, and others figure out how to make real gains with increasingly intelligent, autonomous or semi-autonomous robotic systems. We place the majority of our emphasis on building capacity to do this work in the developing world through the creation of local partnerships called “Flying Labs.” These labs train the next generation of robotics users, implement projects to assist with things like disaster relief and global health, convene discussions around best practices and change in the field, build repair and maintenance abilities, and help to incubate new businesses so that people in the countries where we work can increasingly take the lead.
We got started in this work for a couple of different reasons. One was based on experience in humanitarian aid — just seeing how haphazard some of the deployments were and how much they would have benefited from having strong local partners. So, we committed to help build those local partners. Then, from experience in organizations like Drone Adventures, which is where two of our co-founders came from, seeing the limitations of short-term projects and wanting to do more to sustain activity over the long term. Drones can be a huge benefit to social good work, but only if we’re paying attention to the human context, to regulatory and policy systems, and to sustainability issues rather than just technical wizardry — so that’s what WeRobotics is committed to improving.
The interest in using drones for humanitarian and disaster relief seems to be growing. Why do you think that is?
I think it’s a couple of things — one is simply the dynamic you see with a lot of other technologies, like smartphones, where the market hits a tipping point where costs start to fall pretty dramatically at the same time that usability increases quickly. Drones are also a lot cheaper and better than they were even a couple of years ago. The integration of things like sense-and-avoidance systems into commercial platforms like the DJI Phantom 4 Pro has made flying them much safer. Furthermore, cameras have gotten far better, and software has gotten easier to use without really altering the price much.
Another factor is familiarity. I’ve maintained for a while now that we wouldn’t start to see real gains from drones in the social good sectors until the technology itself matured to the point where it got somewhat boring or at least mundane. We’re almost there. Two years ago, people at conferences would get crazy excited about seeing drones fly. Nowadays, the reaction is more like, “Huh, that’s nice. So now what about the data?” And it’s at THAT point, where we’re more concerned about the data than the hardware, that we’ll start to see projects with real impact.
In the context of lessons learned and doing things better, what do you think is needed to make the drones for social good industry more efficient?
I think one of the things that is needed is building much more local capacity. We’ve started down this path but we still have a LONG way to go. In many parts of the developing world, it is still far too difficult to find places that can offer high-quality repair and maintenance services; manage large volumes of image data; expand the number of folks capable of doing photogrammetry and GIS; and retrieve information that’s truly valuable from a decision standpoint from large imagery datasets. Some of that is going to come through improvements in training — some of it is going to be addressed through machine learning and AI to automate processes — but it just takes a lot of effort in the same direction for a reasonable period to make sure that we’re not always operating on an import model. For drones to have a real long-term impact on the most serious problems confronting the planet, we’re going to need people on the front lines of the most impacted places who are dealing with poverty, disaster, climate change, etc., who are able to use the full spectrum of this capacity, ask the right questions, and implement the solutions that do emerge. We also have to make sure that gains are staying with the people who live in these places, and not simply accruing back to international companies.
When you think of the future of the kind of work you’ve talked about here, what gives you a sense of hope?
What gives me a sense of hope most of all is just working with people in each of the places where WeRobotics has engaged. The sheer excitement, intelligence and novel ideas, particularly of young people, are contagious. When you go to a humanitarian drone training and the entire community comes out to greet you, not because they’re afraid or want to chase you out, but because they really want to see how this stuff works and how they can use it to make their communities better, more resilient, and wealthier — that’s what keeps all of us going, I think.
What are the major challenges from a technology perspective that inhibit more efficient use of drones for social good? Where does more innovation need to occur from a technology standpoint?
Well, the biggest challenge isn’t technological, but rather on the regulatory side. We’re in a period right now where you’re seeing regulations being written to govern drone use all over the world. Some of those regulations are quite good — I would cite the Civil Aviation Administration of the Dominican Republic, for instance, as a positive model — and some aren’t so good, going so far as to threaten prison time for minor infractions like licensing violations. Most often, I see that regulators just need more experience with the technology to understand how it plays a useful role, allowing an activity to happen without violating safety or privacy. It remains an uneven process, and more often than not, it’s a regulatory issue blocking us from working, and not a technology issue.
That said, on the technology side however, I would say that more progress needs to happen on batteries. We are still quite limited by flight times, even for robust platforms. Internet connectivity also remains a bottleneck. Moving imagery files around can be a data-intensive process which taxes network infrastructure. We need better ways to process data locally and to make sure that we can quickly get to the valuable insights contained in drone data.
What advice would you have for a junior practitioner, who might be embarking on a drone career for social good?
I guess I would say that you should approach this kind of work with good humor and with a sense that history or “progress” doesn’t always move in a straight line. Don’t get frustrated or discouraged when it seems like good ideas aren’t gaining enough attention or when funding falls through. Make alliances wherever you can. And always, first and foremost, focus on the beneficiaries of the work. Be a good listener to people in the communities where you’re trying to have a positive impact. That goes for all kinds of “social good” work, but doubly so perhaps for efforts to implement somewhat advanced technologies as part of the toolkit that we’re all using to solve challenging social problems.