3D Robotics

Cross-posted from the 3DR blog.

This past March, Mark Jacobsen (pictured above holding an Ed Ansley Peace Drone prototype), a PhD candidate in Political Science at Stanford, spent a week in Turkey conducting field research among Syrian refugees and activists. At that time the sieges by the Syrian government were at their worst, and the refugees Mark spoke with told him tragic stories, venting their frustration with the U.S. for not airdropping food and other supplies to the besieged areas. Mark, an active-duty C-17 pilot in the U.S. Air Force, explained why the U.S. can’t do airdrops in Syria: Manned cargo planes are vulnerable to anti-aircraft fire, so you can’t conduct airdrops without also launching a major combat operation to take down those air defense systems.

Normally the discussion would end there. That night, though, moved and troubled by the stories he’d heard and the people he’d met, Mark couldn’t sleep. The more thought he gave the issue, the more absurd it seemed that in the 21st century we still don’t have a way to deliver critical humanitarian aid through contested airspace so it gets to those who need it most. Mark wouldn’t let the problem go, and a few months later he launched the Syria Airlift Project.

The project

The Syria Airlift Project (SAP) seeks to end the use of mass starvation and medical deprivation as weapons of warfare. The group and its partners are exploring creative ways to deliver humanitarian aid in conflict zones that are inaccessible to traditional aid organizations. They’ve focused their initial efforts on Syria, where according to a 2014 report by the Office of the High Commissioner for Human Rights, an estimated 240,000 people have been besieged and deprived of both food and medicine, and seven million more are considered difficult to access because of violent and chaotic conditions on the ground.

Mark, however, believes that fleets of small, inexpensive and easy-to-use drones can cheaply and safely deliver critical humanitarian aid—via a large amount of small packets—where larger aircraft cannot. “Imagine an army of ants stealing a picnic lunch, bite by bite,” Mark says. Or perhaps more constructively, an army of people contributing to a massive encyclopedia, byte by byte.

The project aims to use swarming principles and simple airdrop mechanisms to unload bundles of food or medicine outfitted with parachutes at pre-programmed GPS coordinates. The drones are small and quiet and will fly at night so that hostile actors won’t be able to track them; if they do, the drones wouldn’t even be worth shooting down one at a time. Because there are hundreds of them and they’re so small and cheap, no one aircraft is all that important, and it’s Mark’s bet that almost all will complete their mission.

The technology

The whole premise of the Syria Airlift Project is that effective drones can be built cheaply and in large numbers. Made of a simple airframe built from Dollar Tree foam board, packing tape, and hot glue, each aircraft (all-in, with servos, motors, props, autopilot and batteries) costs less than $500. Mark chose to use 3D Robotics APM 2.6 autopilots because they’re inexpensive and widely available, and also because the open-source software can be customized for specific purposes. The SAP’s autopilot modifications include a self-destruct mechanism to prevent the technology from falling into the hands of hostile actors, as well as special navigation instructions in the event of GPS loss or jamming.

The SAP also needed a low-cost way to coordinate simultaneous flights of dozens or even hundreds of these drones, so they created Swarmify, a custom mission planning tool. Mark says that if you give Swarmify one original flight plan, the software can create any number of slightly different flight plans that are randomized by altitude, route and timing. This allows them to quickly create and upload nearly identical and simultaneous flight plans for a whole swarm of drones, while also ensuring safe and collision-free flight.

When a mission comes up, a core team of experts will initiate mission planning and deploy select mobile field crews in Turkey to launch and recover the drones. In flight, the drones will maintain relatively low altitudes as they cross the border into Syria, where they’ll drop their cargo by parachute at the designated coordinates before returning to Turkey. These missions will be coordinated with both the Turkish government and with an extensive network of both Syrians and international aid organizations.

The people

The simplicity of the aircraft offers other advantages, too. Because the drones are built from common materials like foam board and hot glue, refugees need no special skills to take part in assembly, and once trained by visiting instructors, a four-man team can build ten a day. To that end, the Syria Airlift Project also partners with People Demand Change (PDC), a U.S. nonprofit based in Turkey that seeks to empower local populations, who will oversee the employment of Syrian refugees in assembling these airframes.

In this way, Mark hopes his project will also empower Syrian refugee communities, providing meaningful work to those who otherwise have limited means to contribute to their own cause. Aircraft construction will take place in schools and refugee camps in Turkey, where children will have the opportunity to decorate the airframes. The drones will then deliver parcels of food and medicine labeled not only with national and corporate sponsors, but also with symbols and language communicating a shared positive vision of Syria’s future.

The path

Obviously, the challenges the project faces aren’t exactly minimal, among them assessing and minimizing risk of of military retaliation or escalation, preventing unwanted technology transfer, and complying with U.S. and Turkish law to secure a sound legal basis for entering Syria. But the project is still in an early phase. Until now their work has primarily been exploratory, but this weekend Mark will publicly unveil the Syria Airlift Project at a pitch contest hosted by the Defense Entrepreneurs Forum (DEF), a private organization that exists to encourage innovative thinking in the Department of Defense. In the coming months the SAP plans to incorporate as a nonprofit organization, publicize their efforts and begin fundraising.

You could conservatively call Mark’s vision ambitious. At one point he tried to build a drone made entirely of granola, flying food that could be eaten upon crashing. He convinced his wife to cook up a prototype, but they never got the Granola Wing off the ground. (Visit the original 3DR blog post to see a pic.)

“What we’re proposing is insane, I know,” says Mark. “It likely can’t be done. But I keep reminding myself that perhaps it can, and if so, it might save thousands of lives and have a real impact on the Syrian civil war, creating a positive foundation for cooperation on which a shattered society can begin to build.”

We here at 3DR are proud to sponsor Mark and the Syria Airlift Project. If you’d like to help, too, you can contact them at info@syriaairlift.org.

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones


  • Typical food package have about 10-20 kg and more and big size, so there is need develop very big drone for deliver it. Like this or heavier and w much more range of flight:

  • Entirely workable, albeit low impact solution.

    I think this is a workable solution for getting low mass, essential stuff like crucial medicines into cut-off areas.

    And probably not "food" as such, but vitamin pills ?

    A friend and myself have played with quad copters, where we ran the same mission simultaneously, but with a 5 meter vertical separation, looked very cool!

    10  miles is quite doable with cheap airframes, and even with the aircraft retuning.

    I can drop a 250 gram box of pills 10 miles away, and still get my X-UAV Talon back.

    At night, 300 meters up, you won't even know it was there.

    Set up the mission so that it descends to 30m AGL at the target zone, the package should land within a 100m circle, no parachute required!

  • Scott: We have only been using the Ansley Peace Drones for testing, and plan to continue experimenting with llarger aircraft. The APDs can only carry a couple pounds, and we'd eventually like to get 5 lbs or more. We have a couple villages identified that are only about 10 miles from the Turkish border for early testing, so that might be doable. We'll never feed the country with those weights, but we could move critical medical supplies. The nice thing about the project is we're not tied to a specific airframe, so we can continue to iterate and experiment as we go.

    Harry: the name "Granola Gay" was contributed by the article's author, not by me. :-) 

  • I like their original idea of an edible airframe, but with a name like "Granola Gay", it was doomed.  I can see million$ needed to come up with a moldable and edible nutritious material.  Hard candies are moldable, but sugar alone isn't very nutritious and what do you do with all the servos and motors after you eat the airframe?

    Anyway, I see he moved beyond that interesting idea.  I hope he has success in getting food where it's needed. 

  • What kind of payload weight are you aiming for with this design?

  • Developer

    Best of luck to Mark.

    From the technology side, the multi-vehicle support is something that I'm very interested in (primarily for Search and Rescue) and we're making some concrete steps on that.  In particular AC3.2 has a few fixes in it to ensure it behaves properly if it's on the same network as other MAVs (i.e. it checks the set-flightmode message's target before executing it) and you'll also see that MichaelO is building out the mission planner to support multiple vehicles (press Ctrl-X and a drop-down appears with a list of all detected vehicles).

    The RFD900 radios already come with software that allows multiple vehicles to talk on the same telemetry link and that (open source) software can also be loaded onto the 3dr radios.

    Basically anything you can do with one vehicle you can probably do faster with multiple vehicles.

This reply was deleted.