Technology has advanced so far, so quickly that unmanned systems —in the air, on land, on or under the sea —will soon have the capability to undertake missions that we have yet to fully define or understand.

By late 2020 for example, the U.S. Air Force could have fighter aircraft-sized UAS that could fly in formation with manned F-35s or F-15s, break off to reconnoiter 100 miles away, and launch their own smaller drones to strike targets on the ground.

While that capability may be just a couple years away, it will shortly be augmented by others. The Army is a partner in DARPA research into unmanned air and ground vehicle swarms. In its 2020 defense budget, the U.S. Navy increased research and development spending on unmanned underwater vehicles (UUVs) a noteworthy 160 percent.

But when the hardware materializes, who's going to control it and how? Could one pilot, one Soldier, or one Sailor really take on the task of controlling one, or even up to 250 drones?
 
Loyal wingmen
 
A few months ago, a 29-foot-long UAS with a 27-foot wingspan took off from a runway at the Army's Yuma Proving Grounds in Arizona. It looks like a stealth fighter, albeit one without a cockpit. The aircraft is Kratos Unmanned Aerial Systems' XQ-58A Valkyrie demonstrator, which completed a 76-minute first flight on March 5. Valkyrie is a high-subsonic strike aircraft which can climb to around 50,000 feet and has an unrefueled range greater than 3000 nautical miles, which gives it about 10 hours' worth of endurance.

The XQ-58 is part of a joint effort between Kratos and the Air Force Research Laboratory (AFRL) under the latter's Low Cost Attritable Aircraft Technology (LCAAT) program. Boeing is developing a similar unmanned combat air vehicle (UCAV) known as Loyal Wingman. Both UCAVs are designed to be relatively cheap (potentially $1.5-$2.5 million each) and fit into existing tactical aircraft formations and command structures.

The XQ-58's stealth shape makes it capable of accompanying manned penetrating strike aircraft like the F-35. At their direction —or command from a combat systems officer on E-8 JSTARS aircraft or an Air Force combat controller in the ground —the XQ-58 is designed to maneuver to a target, deploying Small Diameter Bombs from its internal bomb bay.

In May, Kratos announced it was teaming with drone maker Aerovironment to integrate a small tube-launched missile called Switchblade with XQ-58. Though it's called a missile, Switchblade flies like a small UAS until the point where it hits its target and explodes. As such, it's also known as a kamikaze or suicide drone. With a limited ability to loiter, Switchblade is capable of relaying information back to the Kratos mothership as well as striking targets.

The combination would give an F-35, accompanied by several XQ-58-type wingmen, the option to assign different drone wingmen to pass sensor data to other manned/unmanned aircraft, to surveil a target, or execute a strike.     

But fighter pilots, airborne weapons officers and ground controllers say they're already task-saturated managing their manned assets. How could they add multiple large or small combat drones? With a tablet-like control interface which allows them to simply direct highly autonomous drones, says Steve Fendley, president of Kratos Unmanned Systems.  

"We've developed our tactical systems including Valkyrie, to enable a tablet to issue command and control [instructions] at a very high level," he says. "The tablet enables one operator to control many unmanned aircraft. They can be different types of unmanned aircraft. You can coordinate the missions between them, you can task them independently, you can task them as groups. You can re-task them."

The summation possibly suggests a tablet for Valkyrie or other drones with simple virtual button commands like "rendezvous," "reconnoiter coordinates X-Y" or "strike target Z."

According to Fendley, a tablet operator "can certainly be doing something else that is taking most of his attention."

If so, the question remains as to whether a fighter pilot or combat controller would have a good enough battlespace picture to effectively orchestrate drone wingmen.

Kratos' self-funded development program has already put sufficient autonomy in place to allow XQ-58 and smaller drones, such as the company's MQM-178 Firejet, to independently undertake basic maneuvering, navigation and aircraft avoidance. The Valkyrie can maintain itself in a "safe situation" maneuvering tactically with other drones/fighters without the operator having to intervene, potentially right up to weapons release which requires human-based approval.

Despite its early development, Valkyrie is already raising possibilities that are ahead of the scope of current Air Force operational schemes and operators Fendley acknowledges.    

"Especially with respect to the mix of manned/unmanned operations together. Those techniques, tactics and rules sets are all within the early stages of their development."  

A complimentary AFRL program called "Skyborg" is looking at tactics development. It posits a "dialable" artificial intelligence for drone wingmen. A pilot would decide just how autonomous a Valkyrie or other drone should be on a given mission.

Because of the unknowns, Kratos/AFRL hope to clarify the operational employment of unmanned strike/ISR aircraft by experimenting with smaller drones even as XQ-58 proceeds with testing.  
 
Urban swarms  
 
Small drones and small units are the focus of DARPA's OFFSET (OFFensive Swarm-Enabled Tactics) program. It looks at the tactical possibilities of small ground units employing mixed UAS/UGV swarms for situational awareness, clearing buildings and other tasks in urban settings where high vertical structures, tight spaces, and limited lines of sight constrain military communications and tactics.

DARPA program manager Dr. Timothy Chung characterizes OFFSET as a discovery program envisioning the use of air/ground swarms of 250 unmanned vehicles or more, controlled by one or two individuals within or remote from a small infantry unit. OFFSET is "Providing a vision of what could be, should these technologies mature," Chung says.

The concept raises many questions; What types of drones and ground robots might be used, with what sensors? How much attrition could a swarm withstand before becoming ineffective? How could units with depleted swarms be resupplied or rotate out of action? Would Army/Marine swarms be interoperable and what tactics would small units use with swarms?

These are issues for the services to define Dr. Chung asserts but OFFSET is looking at them. That includes the question how one or two individuals could control a swarm of 200-300 drones.

"We're exploring how to make dramatic advances in how to interact with swarms, not only how a human commander delivers commands to the swarm but also how the swarm processes information and conveys information back to the commander."   

Chung suggests an interface device incorporating technologies like augmented virtual reality, haptic/gesture commands, potentially even a digital assistant working alongside the human. What sort of individuals could grapple with the data and tactics to effectively employ swarm?

OFFSET has explored what it calls three different "personas" for the control of swarms.

A "Swarm Mission Planner" might reside behind the front lines, controlling swarms actively or temporally, having a conceived pre-mission plan. A "Swarm Tactician Rear" could be near the operational scene but not in immediate combat proximity, using different swarm sensors to orchestrate different tactics. A "Swarm Technician Forward" embedded with the unit would manage swarm behavior and tactics in the midst of battle.

Chung acknowledges that "no one interface is going to allow you to transcend all three personas."

It's not by accident that the three personas are roughly analogous to a tactical commander at battalion level, a pilot or tactical operator loitering in a fighter or battlefield control aircraft, and a forward air controller with a unit on the ground.

"The vision of OFFSET," Chung adds, "is to use the swarm as just another tool in your inventory. When I think of swarms, I would say it's not just a collection of systems. Rather the swarm is a system itself."

The swarm-as-a-system approach is perhaps an easier way for swarm controllers to get their heads around combat operations. While the capacity to create large swarms exists, operational concepts and necessary artificial intelligence parameters don't. That's OK, Chung says. 

I think the vision of OFFSET and its heterogeneous forms helps us learn about what to use, when and where ... especially in the context of technology turnover they fall into development for the future."

Raytheon's Barracuda could work with other systems to help take out mines. Image: Raytheon

Undersea mine neutralizers
 
The U.S. Navy is in the early phases of automating undersea mine detection and neutralization. By combining search/identify and neutralize elements on a single unmanned platform executing a single sortie, mine detection and neutralization could be conducted in a single pass of an area, saving vital time and ship resources. The concept is called Single Sortie Detect-to-Engage (SSDTE).

A Navy Littoral Combat Ship (LCS) would launch the service's Common Unmanned Surface Vehicle (CUSV). On reaching its patrol area, the CUSV would deploy a sophisticated sonar that can spot mines from any angles in high definition. If an object is determined to be a mine, it could be detonated by another drone which would launch from the CUSV.

Raytheon's Barracuda is a sonobuoy-sized unmanned, underwater vehicle that can find and destroy near-surface, volume and bottom sea mines autonomously. When the CUSV's sonar detects a mine, the USV operator gives it a command to launch Barracuda which is ejected from an onboard sonobuoy tube.

The UUV enters the water, then separates from its own surface communications buoy and begins seeking the mine located by the CUSV using its own sonar and electro-optical sensors. As it does so, it communicates with the surface buoy which relays its data/underwater pictures to the CUSV which forwards them to the LCS.

How would Barracuda be controlled and by whom? That has yet to be fully determined, Raytheon's Deputy Program Manager, Dave Chapman admits. "There are many aspects of the emerging autonomous missions associated with neutralization as well as mine hunting mission that are currently being evaluated," he says.

The company envisions separate CUSV and Barracuda operators, allowing that multiple USVs and UUVs may be mine hunting at the same time. Whether the Barracuda operator is a mine warfare specialist or another sort of specialist isn't yet known. The control interface isn't yet fully defined, either, though it will be separate from the CUSV. Clearly, the quarterbacks and plays will have to be figured out for SSDTE missions.

High-level vehicle autonomy will be key. Barracuda seeks and identifies mines then loiters nearby while awaiting confirmation and kill authorization from its shipboard operator completely autonomously.  

"The neutralizer is able to perform all aspects of the mission," Chapman says, "without any operator intervention or control up to the point of initiating the destructive charge. The operator supervisory function is necessary to ensure any hazards associated with the neutralizer warhead itself or the ultimate detonation of the mine are able to be mitigated."

Once a neutralization authorization is given, Barracuda detonates in proximity to the mine, blowing itself and the mine up. Raytheon maintains that it's a cost effective solution but it would appear that more than one quarterback would be needed to make an SSDTE mission —with two drones or 10 —happen.   

It's a reminder that a play often takes an offensive coordinator and other team players. Can one individual oversee dozens of autonomous unmanned vehicles? We'll soon find out.

Above: An artist's conception of a drone swarm. Below: Kratos' Unmanned Aerial Systems' XQ-58A Valkyrie. Photo: Kratos