The swarm is coming. Packs of centrally monitored, flying robots are now a hallmark of the Pentagon’s strategy of future warfare.

By leveraging advances in robotics, big data, 3D printing and miniaturisation, the solo unmanned air vehicle will be replaced by the swarm – a dense network of collaborating, expendable UAVs that autonomously pool their resources to perform a mission.

The original concept dates back more than a decade, when the US Defense Advanced Research Projects Agency tried to develop the heterogeneous airborne reconnaissance team, but the idea was still too far ahead of its time. A series of successful demonstrations performed in 2016 – along with the outspoken support of top Pentagon leaders – suggests the swarming concept is rapidly becoming a practical reality.

Leap forward

Like so many innovative concepts, shifting swarming from the laboratory to an operational environment is not going to be straightforward. That step requires an even more radical change in the industrial world. The military must find suppliers that can deliver sophisticated payloads, military-hardened electronics and flexible airframes, yet somehow at a cost an order of magnitude lower than the typical small unmanned air system. The supplier must be innovative with technology and business models, yet intimately familiar with the Pentagon’s acquisition bureaucracy.

Among a host of Silicon Valley-backed start-ups and university research teams, Raytheon Advanced Missile Systems (AMS) is positioning itself – and the 5.9kg (13lb) Coyote unmanned air system – as the optimal solution.

The AMS unit may not seem like the obvious place to start looking for a low-cost expendable UAS provider. Its engineers are well-known for developing the next generation of missile defence interceptors and air-launched weapons, each costing in the order of $1 million with every shot. But Raytheon AMS came into the swarming business through the acquisition in 2015 of a Tucson neighbour.

Advanced Ceramics Research (ACR) was founded in Tucson in 1989 to develop ways to produce the high-temperature material embedded in the company’s name. The goal was to develop new ceramics to use in drill bits for the mining industry. But the company’s founders happened to include model airplane enthusiasts, which led to an interest in the burgeoning unmanned air vehicle industry of the late 1990s.

Coyote UAV Raytheon

During collaborative flight demonstrations performed last year, Raytheon showed that its low-cost Coyote could transform reconnaissance


ACR developed hand-launched Silver Fox and Manta UAVs, which had no need for ceramics but gained the company exposure with the Pentagon. An approach from the US Navy led to the development of the Coyote, a UAV that could be launched from the sonobuoy tubes in Lockheed Martin P-3C Orion maritime patrol aircraft and Sikorsky MH-60R Seahawk helicopters.

That design requirement for a small UAV that can survive a launch tube ejection in flight gave the Coyote a uniquely adaptive capability. If it can be air launched, the Coyote also can be tube launched from the ground or aboard ships.

“It is not trivial where you can drop these things out of a tube on an airplane and make it fly, nor is it shooting out of a launcher and make it fly. There’s a lot of people who have been working on this for a long time,” says John Hobday, business development lead for Raytheon.

By 2009, the Coyote programme helped make ACR an acquisition target. BAE Systems acquired the company, helping the former mining materials supplier navigate the Pentagon’s labyrinthine acquisition system. By then, the Coyote had also gained interest as a weather reconnaissance system for the US National Oceanic and Atmospheric Association. Six years later, Raytheon acquired ACR from BAE Systems.

Fresh approach

By then, the Pentagon was rolling out a new strategy called the Third Offset, which sought new ways of delivering advanced capabilities without breaking the bank. The formation of the Strategic Capabilities Office inside the Pentagon, which was not officially acknowledged until a year later, immediately focused on the possibilities of swarming UAVs.

The Office of Naval Research had also taken an interest. A series of simulations by naval experts had reached a disturbing conclusion: swarms of small boats attacking an Aegis-class destroyer in a port environment would often be successful. The Aegis combat system was adept at thwarting advanced attacks by high-flying fighters armed with supersonic, sea-skimming anti-ship missiles, but less equipped to defeat a horde of small boats laden with explosives. To overcome this new, low-cost threat, the navy began looking for a low-cost solution.

Thus, the ONR’s Locust programme was born, with Coyote as the selected platform. Until Locust, the Coyote had been operated only as a single-ship asset, performing missions solo. To be adapted to the swarm mission, the Coyote needed a few upgrades. First, it required new software that centrally manages the swarm, assigning roles to specific aircraft in the network, then reassigning roles if priorities or circumstances change. Second, the Coyote needed a launcher that could dispense 30 aircraft within 1min.

In two demonstrations staged last year, the Coyote proved that it could be launched as a swarm and operate collaboratively, says Hobday.

“They were very successful. Everything went exactly as planned. We met all objectives,” he says.

The demonstrations showed how a swarm of Coyote UAVs can change how the reconnaissance mission is performed. When a single UAV or manned reconnaissance aircraft is used to observe a target today, their sensors are unable to see behind walls and corners. The swarm approach solves that problem. Multiple UAVs can track the target from different angles, leaving no place to hide. In other scenarios, a low-cost swarm can be used to overwhelm an enemy’s defences, flooding the airspace with too many UAVs for them to shoot down.

First, however, the UAVs have to be truly expendable. Unmanned systems deployed by the military today are intended to be recovered after every mission, as their sensors and electronic components are too expensive to replace.

The US Air Force Special Operations Command has commissioned Raytheon to develop a version of the Coyote that could be launched from a Lockheed AC-130 gunship. Instead of exposing the AC-130 to ground fire in low passes, the gunship could launch a swarm of Coyotes to investigate the target and transmit video back to the mothership. In that scenario, the Coyotes would not be recovered. The controller on board the AC-130 would command the Coyotes to dive into the ground after completing their mission.

To afford hundreds of thousands of expendable, flying sensors, the US military needs a UAV sold at commodity prices. The wiring and circuit cards would still have to be military-grade: after all, the military cannot allow an enemy to remotely jam or take control of the Coyote swarm. The airframe must also be strong enough to survive the tube launch from either the ground or the air.

In phase 2 of the Locust programme, the ONR plans to have Raytheon demonstrate that it can solve this manufacturing problem. For swarming to work, Raytheon has to be able to build hundreds of thousands of Coyotes at a low enough price that the US military can throw them away after each mission. The company’s engineers are already working on making changes. The Coyote’s composite wings, for example, consist of dozens of plies that are laid-up by hand. Raytheon wants to switch to a mould injection system, yet preserve the strength characteristics of the wing, Hobday says. The next step is to develop a way to manufacture the military-grade circuit boards and wiring.

“It’s working with those suppliers and pushing the cost down on them, because first off you’re manufacturing a lot of them,” Hobday says. “And it’s changing the way that you look at some of those circuit boards and components.”