The Committee on Counter-Unmanned Aircraft System (CUAS) Capability for Battalion-and-Below Operations recognizes that the U.S. Army and other Department of Defense (DoD) organizations are engaged in a significant number of very relevant and often highly successful materiel and non-materiel (e.g., tactics and training) efforts to address the single-small unmanned aircraft system (sUAS) threat. The committee thus did not see a need to discuss these efforts in detail in a report that is focused on science and technology efforts. Some specific areas of concern (e.g., the inability of a dismounted infantry unit to detect and engage a surveillance UAS at a distance of a kilometer or more) relate to single sUASs but are also of concern for large groups of sUASs. However, this report is mostly focused on the future threat of coordinated groups, swarms, collaborative groups, and collaborative swarms, defined later in this report, because (1) the committee anticipates these threats will appear much earlier than does the Army, (2) few development efforts appear to be focused on countering these threats, and (3) the sponsor, the Assistant Secretary of the Army (Acquisition, Logistics, and Technology), asked that the committee focus on collaborative groups, as discussed below in “Study Origins.”
The development of the statement of task began in January 2016. At that time, concerns about the adversarial use of commercial sUASs was emerging. Since then, a significant number of media reports have discussed both state and non-state actor development and use of sUASs for military operations. These activities figured prominently in committee deliberations, and some are discussed in subsequent chapters.
This study’s origins were based on concerns due to the confluence of multiple developments. First, the worldwide availability of relatively inexpensive and significantly advanced sUASs, especially small hobby aircraft, created opportunities for potential adversaries to easily acquire sUASs with impressive and rapidly growing capabilities. Second, these readily available, high-performance sUASs pose a significant potential threat to U.S. forces as lethal weapon systems; reconnaissance, surveillance, and target acquisition (RSTA) systems; and electronic warfare systems (especially as low-power jammers). Lethal sUASs can carry either conventional (kinetic or non-kinetic payloads) or unconventional (chemical, biological, and radiological) payloads. Third, the size, weight, and power (SWaP) needs for most developed or developing counter-sUAS systems make them more appropriate for use on medium and heavy vehicle platforms or for static emplacement in the defense of fixed sites. Nearly all of the current counter-sUAS systems do not fit within the available SWaP of an infantry unit. After discussing these concerns with senior leaders within the U.S. Army Training and Doctrine Command Headquarters, and three of its subordinate organizations, the Fires Center of Excellence
(COE), Maneuver COE, and Cyber COE, it was agreed that a study should go forward and that it should focus not just on dismounted infantry, but also lightly armored vehicles in battalion-and-below operations. These latter two points are particularly important because dedicated air defense units within brigade combat teams were withdrawn from Army inventory in the early 2000s after a determination was made that no significant air threat existed to maneuver battalions and lower-echelon units due to the demise of the Soviet Union.
The commercial market for small hobby aircraft has increased dramatically over the past 3 years. Venture capital firm Kleiner, Perkins, Caufield, and Byers has estimated that global sUAS shipments reached about 4.3 million units in 2015, up from 1.6 million units in 2014. About 35 percent of those sales were in the United States (Bedard, 2015). Within the United States, the number of sUASs sold from April 2015 to April 2016 grew 224 percent. More disturbing is that Chinese manufacturer DJI accounts for 50 percent of the sUAS sales in North America, while 3D Robotics (American) and Yuneec (also Chinese) trail with 7 percent and 4 percent of the market share, respectively (French, 2016). This surge in sales motivates manufacturers of hobby sUASs to make substantial investments into improvements in sUAS capabilities such as increased payload capacity, advanced aerodynamic performance, autonomous navigation, advanced avionics, and the use of smartphone and/or tablet systems as operator control stations. Many of these advances are developed by non-U.S. companies and personnel. As an example, the DJI Phantom 4 has enhanced flight modes that use Global Positioning System (GPS) Waypoints (i.e., it flies along operator-selected GPS points), Home Lock (i.e., the user identifies a GPS-set position as “home” to which the Phantom 4 returns when it has problems), and Point of Interest (i.e., the sUAS flies in circles around an object or location).1 The flight modes include the following capabilities:
- Obstacle avoidance. Avoid obstacles, using an array of ultrasonic rangefinders (aiming left, right, forward, backward, and down), vision sensing (visual cameras integrated with computer vision algorithms), and onboard processors. This mode does not require GPS.
- Tapfly. Tap on a point in the video image (from the Phantom 4’ s camera) on the control screen (e.g., a smartphone) and the Phantom 4 will fly there.
- Follow me. The Phantom 4 is linked to a mobile device. Both must have strong GPS signals. The Phantom 4 will follow a person, similar to the skier in Figure 1.
- Active track. Track subjects automatically (without using a GPS reference) by using advanced image recognition algorithms. The user can also use it to effortlessly orbit around subjects.
- Sport mode. The Phantom 4 can increase speed to 20 m/s and ascend or descend at speeds of 9 m/s and 4 m/s, respectively.
In addition, flight time is also improving. Previous batteries lasted for approximately 15 min in regular flight mode, but with its 5,230 mAh battery, the Phantom 4 can operate a little under a half hour in regular flying mode. Additionally, prices have fallen just as dramatically as sales have risen.
The bottom line is that consumer sUASs are easy to buy, their performance is improving dramatically, their cost has dropped significantly, and there are millions of them around the world. Thus, they pose a significant and growing threat to U.S. warfighting forces when used for nefarious means rather than as intended.
With creative thinking and engineering, at which current and potential adversaries have proven adept, hobby sUASs can be used to pose a significant threat to U.S. warfighters, especially lightly armored vehicles and dismounted infantry in the U.S. Army, U.S. Marine Corps (USMC), and Special Operations Forces. Today, U.S. military personnel have to be concerned about the use of sUASs by state (e.g., peer nations) and non-state (e.g., Islamic State in Iraq and Syria [ISIS]) actors (Watson, 2017). As documented in various new stories, they may also pose a threat to U.S. Navy ships and submarines while in port and U.S. Air Force planes on airfields and during take-off and landing.
sUASs can be used to support conventional and unconventional attacks. sUASs can be fitted with an external or embedded explosive designed to explode on impact or when handled. As an example of this latter case, on October 2, 2016, two Kurdish soldiers were killed and two French soldiers were wounded near Mosul, Iraq, when an ISIS booby-trapped fixed-wing sUAS exploded (Atherton, 2016). Contrary to the past, when U.S. warfighters may have found improvised explosive devices, now the improvised explosive devices will find our warfighters. There have also been reports of sUASs with attached small arms (e.g., pistols) and even flamethrowers. Because of weight and velocity, a sUAS itself can generate enough kinetic energy to cause significant damage when it crashes into a warfighter or soft target. sUASs can also be equipped with external payloads that can be used to drop explosives, as the ISIS has done in Syria and Iraq2 or, in an unconventional attack mode, for disseminating biological or chemical agents from an altitude that minimizes detection and direct engagement while maximizing contamination and kill ratios.
In addition to the above kinetic kill applications, sUASs can be used by the enemy for RSTA operations, for jamming radio frequency signals, and for supporting information operations. As an example of the RSTA use of sUASs, the Russians have employed as many as 16 different types of UASs in Ukraine (Tucker, 2015). They flew sUASs in pairs and used them with great effect to target Ukrainian units. The sUASs first found Ukrainian units, then operators passed the information to multiple indirect fire systems (howitzers, missiles, and rockets), which delivered massed fires for 3 min. From beginning to end, the process took about 15 min (Freedberg, 2015). This approach has resulted in Ukrainian casualty rates as high as 85 percent (Karber, 2015). Discussions of the threat of sUASs, especially hobbyist sUASs, have been reported in an online Wired article (Barrett, 2016) and in a report (Remote Control Project, 2016) by Open Briefing, a nonprofit think tank in the United Kingdom. sUASs can be fitted with low-power jammers to support localized jamming missions. An example of sUASs in an information operations mission is the employment of a swarm of sUASs to form the flag
of a terrorist organization above a village or town.
Recently marketed sUASs have technological enhancements (e.g., obstacle avoidance and target-following technologies) that support autonomous flying with no need for a control link or access to GPS. These autonomously guided sUASs can overcome counter-sUAS systems based on jamming radio frequency and GPS signals. The threat these capabilities pose are not looming in the future; they are here today. Commercially available, small, autonomously guided aircraft components and systems are already on the market, and more capable components and systems continue to be advertised to consumers. For example, DJI, Intel, Parrot, Neurala, and LeddarTech have developed novel “obstacle avoidance,” as opposed to just “airborne collision avoidance,” navigation technologies using barometers, ultrasonic altimeters, optical flow sensors, camera-based perception, and simultaneous localization and mapping to sense and avoid obstacles. With these technologies, sUASs can accomplish a mission with minimal use of data links or access to GPS signals. Current sUASs sense and avoid technologies are not perfect, but they do offer many amazing capabilities both outdoors and indoors. Most likely, new approaches will be announced even while this report is in development.3
As the statement of task for this study was being developed, and sUAS capabilities continued to improve at a lightning pace, a new concern began to emerge: the threat of swarms and collaborative groups of sUASs and how they may be countered. The statement of task was modified to address this concern before the study was approved. As the study progressed, it became apparent that this concern was based on current, real threats, as evidenced by various news reports and as discussed by invited speakers during unclassified and classified presentations.
For years, the United States appeared to have a clear lead when it came to swarming sUASs. Last year the Advanced Robotic Systems Engineering Laboratory claimed a new world record by launching a swarm of 50 sUASs, all controlled by the same operator.5 More recently, the DoD Strategic Capabilities Office, partnering with Naval Air Systems Command, successfully demonstrated a large (>100) sUAS swarm at China Lake, California. The demonstration, conducted in October 2016 and documented on the CBS News program 60 Minutes on January 8, 2017, consisted of 103 Perdix6 sUASs launched from three F/A-18 Super Hornets.7 The sUASs demonstrated basic swarm behaviors such as adaptive formation flying. Current goals include demonstrating advanced swarm behaviors, such as collective decision-making and self-healing, as well as flying Perdix sUASs in batches of up to 1,000 (60 Minutes, 2017).
However, peer nations are also making progress. At the 11th China International Aviation and Aerospace Exhibition, state-owned China Electronics Technology Group Corporation claimed to have flown a swarm of 67 sUASs (Rambling, 2016). A continued U.S. technological lead cannot be assumed.
The threat of swarms and large collaborative groups pose a significant challenge for counter-sUAS efforts. Not only is it difficult to detect, identify, and track numerous sUASs flying from many
3 Sense and Avoid for Drones is No Easy Feat, http://droneanalyst.com/2016/09/22/sense-and-avoid-for-drones-is-no-easy-feat/.
4 Swarms and Collaborative Groups are defined in detail in this chapter in the section entitled “Modes of sUAS Utilization.”
5 Raymond Buettner, associate professor, Naval Postgraduate School, “State of the Art and Vision for Class 1 VAS Swarming and Collaborative Capabilities,” presentation to the committee on November 7, 2016.
6 This is not an acronym; it is the name of a small bird, a partridge. In Greek mythology, Athena turned Perdix into a partridge to save him while he was falling from a high tower , https://en.wikipedia.org/wiki/Perdix_(mythology).
7 The Perdix UAS was developed by MIT Lincoln Labs.
distributed directions, counter-sUAS systems will need very deep magazines and sufficient stowed kills to neutralize large numbers of threat sUASs (“deep magazine” and “stowed kill” are defined in Appendix C).
Both the Training and Doctrine Command Headquarters and the Fires COE requested that this study focus on battalion-and-below operations to address both the lack of dedicated air defense capabilities within battalions as well as the vulnerability of dismounted infantry and lightly armored vehicles to sUASs. In fact, at the time this report was being written, there were no air defense units in brigade combat teams. Current air defense units are located at echelons above corps and are not resourced to support brigade-and-below maneuver forces on a regular basis. As a result, the statement of task was modified to accommodate the request to focus on battalion-and-below operations.
For this study, the committee assumed an adversary will use sUASs against unit personnel, equipment, and activities within a dismounted infantry battalion in either open and natural or urban terrain. A dismounted infantry battalion was selected because dismounted infantrymen have minimal body armor and have very limited SWaP capacity available to support organic counter-sUAS systems. In addition, the sponsors for this study were most interested in the ability of dismounted infantrymen and lightly armored vehicles to protect themselves from threat sUASs. An extremely large number of potential targets for sUAS attacks are possible when considering the Modification Table of Organization and Equipment for battalions within an infantry brigade combat team along with numerous offensive, defensive, and security scenarios. To make its task manageable, the committee considered a smaller set of unit personnel, equipment, and activities in its deliberations, including the following:
- Command and control elements of the battalion and company;
- Dismounted infantry unit (company/platoon/squad)
- Moving in tactical formations (e.g., column, vee, or wedge),
- Assaulting an objective, or
- Defending a piece of terrain;
- Lightly armored vehicles
- Moving in a column,
- Composing a temporary, static command and control node, or
- Resupply battalion assets in a temporary, static location;
- Battalion or company mortars, such as
- U.S. Army battalion mortars setting up or already set up at a firing position or
- USMC battalion and company mortars setting up or already set up at a firing position; and
- A forward operating base (in both open terrain and an urban environment).
Besides the U.S. Army, the USMC has a significant number of dismounted infantrymen just as vulnerable to threat sUASs. To address this problem, the USMC is sponsoring multiple ongoing technology development efforts, as well as experiments to assess possible materiel and non-material approaches to countering sUASs. The Marine Corps Warfighting Laboratory is one of the focus points for
these efforts. Marine Corps Warfighting Laboratory personnel were approached about the study, and they agreed that the USMC could benefit from the study’s findings and recommendations.
Although not officially approached about the study, the U.S. Navy, the U.S. Air Force, and the U.S. Special Operations Command may also benefit from this study, based on recent reports of counter-sUAS activities in these organizations. For example, the Naval Research Laboratory discussed a recent Joint Emerging Operational Need document for countering sUASs. One of the motivators for this Joint Emerging Operational Need document was a sUAS incident at the U.S. Navy’s submarine base in Bangor, Washington.8 The U.S. Air Force is also worried about drone threats to nuclear sites (Copp, 2016).
Finally, the Department of Homeland Security (DHS) can benefit from this study. Besides its current challenges with sUASs for customs (e.g., transporting contraband across borders), prisons (e.g., flying contraband over prison walls), forest fires (e.g., interfering with aerial firefighting), airports (e.g., threatening take-offs and landings), and sensitive areas (e.g., entering White House area or getting too close to sensitive infrastructure), DHS may see sUASs being used against emergency responders. If true, the emergency responders may benefit from counter-sUAS approaches recommended for dismounted infantry and lightly armored vehicles. Thus, a consideration of what counter-sUAS approaches might be useful to DHS was included in the statement of task.
The Army uses the following timeframes for its efforts to develop force capabilities and for planning research and development efforts: near-term (today-2025, the Current Force), mid-term (2026-2035, the Interim Force), and far-term (2036-2050, the Future Force).9 Additionally, the committee found conflicting Army definitions for the near-, mid-, and far terms across various organizations and documents. For example, the mid-term in the U.S. Army Robotic and Autonomous Systems Strategy is 2021-2030, not the 2026-2035 timeframe mentioned above (MASD ARCIC, 2017).
Based on an open source review of current and anticipated threat uses of sUASs, discussions with the Army about its timeframe and planned development and deployment of counter-sUAS systems, and the tremendous improvement in sUAS performance capabilities in short periods of time, the committee believes that the Army timeframes are significantly out of sync with the rapidly advancing performance capabilities of individual sUASs and teams of sUASs. Additionally, significant sUAS performance enhancements are occurring so quickly that it is impossible to predict performance capabilities beyond 8 years (the length of the Army’s “near term”). Unless potential threat sUAS capabilities and counter-sUAS efforts are addressed more rapidly, the vulnerabilities of dismounted infantry and lightly armored vehicles to sUAS threats will grow extremely quickly, potentially to the point where force protection standards cannot be met for soldiers in the field.
To reflect the speed of sUASs developments that it expects, the committee has generated its own timeframe taxonomy, as follows:
- Immediate (today-2019, 1 to 2 years);
- Imminent (2020-2022, 3 to 5 years); and
- Emerging (2023-2025, 6 to 8 years).
8 John Lee, senior scientist (ST), Applied Optics Branch, U.S. Naval Research Laboratory, “Current and Future U.S. Navy CUAS-related Science and Technology Efforts,” presentation to the committee on December 19, 2016.
9 COL Frank Brewster, (MBL) TCM SBCT, Maneuver Center of Excellence, “MCoE “Counter- Small UAS” Efforts Related to Dismounted Infantry and Lightly Armored Vehicles,” presentation to the committee on March 14, 2017.
The committee uses this taxonomy to discuss both predictions of sUAS performance capabilities and in recommending S&T investments in counter-sUAS systems.
As will be seen in this report, this difference in timeframes creates mismatches between committee predictions and Army predictions about the performance capabilities of sUASs—for example, in the anticipated timeframes for the development of swarm and collaborative group capabilities for sUASs.
Developing effective countermeasures to highly modified and customized sUASs is a difficult challenge. Additionally, a counter-sUAS system often costs significantly more, in terms of per-system and/or per-engagement cost, than the cost of the individual sUASs being countered.
Countering sUASs requires the detection, identification, and neutralization of threat sUASs. Detection and identification are very difficult because these sUASs are small, can fly at low altitude, and can have highly irregular flight paths that can range in speed from zero (hover) to close to 18 m/s.10 Additionally sUASs can take advantage of the significant amount of background clutter close to the ground (e.g., birds and trees).
Once detected and identified, neutralizing a sUAS is a separate and even greater challenge. For neutralization, the DoD has been developing various defenses against sUASs, but mostly for use against individual sUASs and especially those larger than hobby aircraft. Kinetic counters, such as shooting down a single, highly dynamic, fast-moving, low-flying hobby aircraft with small arms (rifles, shotguns, and light machine guns), are extremely difficult due to the agility and small size of sUASs. Additionally, swarming sUASs can be employed to overwhelm most existing kinetic countermeasures. Finally, any counter-sUAS system developed for dismounted infantry, who are already overburdened with equipment, must minimize the additional SWaP and cognitive demands on the infantryman.
To assist in identifying sUAS technologies and how they can contribute to future sUAS capabilities, the committee decomposed sUAS functions and capabilities. The decomposition includes the following four areas:
- Autonomous behavior,
- Supporting functions,
- Mission packages, and
- Development and testing needs for sUASs.
Each area is further decomposed into more detailed subareas. This is discussed in Appendix D.
One challenge the committee encountered was the variability of definitions across organizations. In this section, the committee presents the definitions it used in conducting its work and authoring this report.
10 Specifically for multi-rotor aircraft. Fixed-wing sUAS and hybrid sUAS (able to hover as multi-rotor sUASs and rotate the whole aircraft to transition into a fixed-wing mode) can achieve much higher speeds.
TABLE 1 The Five Department of Defense Groups of Unmanned Aircraft Systems (UASs)
|UAS Group||Maximum Weight (lb.)a||Nominal Operating Altitude (ft.)||Speed (mph)||Representative UAS|
|Group 1||0-20||<1,200b||<115||Most commercial hobby drones (e.g., DJI Phantom series), RQ-11 Raven, Wasp|
|Group 2||21-55||<3,500b||<290||ScanEagle, Silver Fox, BUSTER|
|Group 3||<1,320||<18,000c||<290||RQ-7 Shadow, RQ-21 Blackjack, RQ-23 Tiger Shark|
|Group 4||>1,320||<18,000c||Any airspeed||MQ-8 Fire Scout, Predator, MQ-1C Gray Eagle|
|Group 5||>1,320||>18,000c||Any airspeed||MQ-9 Reaper, RQ-4 Global Hawk, MQ-4C Triton|
NOTE: The units of pounds (lb), feet (ft), and miles per hour (mph) are retained in this table to match Department of Defense definitions.
a Maximum takeoff weight.
b Above ground level.
c Above sea level.
SOURCE: Adapted from https://en.wikipedia.org/wiki/U.S._military_UAS_groups.
Unmanned aircraft systems (UASs). This is an aircraft without a human pilot aboard. UASs were previously referred to as unmanned aerial vehicles (i.e., UAVs) within DoD and are often identified as drones in DoD and, more commonly, in the media.
Class (a.k.a. group) 1 and 2 UASs. Within DoD, UASs are identified in five groups, shown in Table 1. This report mostly focuses on sUASs that are assigned to Group 1, especially small hobby aircraft, which are described in more detail immediately below Table 1 (DoD, 2011). This report uses the term “class” instead of “group” to avoid confusion with discussions later in the report that use “group” in reference to multiple sUASs.
As shown in Table 1, sUASs weigh up to 55 , lb. (25 kg). Included in the sUAS category are most commercial hobby aircraft (e.g., quadcopters), which weigh less than 5 lb. (2.3 kg) and have a diameter less than 2 ft. (60 cm). A hobby rotary-wing aircraft of this size can have an average speed of 18 m/s, fly at altitudes ranging from 50 cm above the ground to 6,000 m above sea level,11 hover vertically with a precision of ±8 cm, fly more than 20 minutes with a standard battery, and easily carry payloads weighing 1 kg. Designs vary significantly, as discussed below, and there are many variations of these capabilities—for example, higher speeds, longer flying times, and heavier payloads for fixed-wing aircraft.
Types of sUASs. The types of sUASs are fixed wing, rotary wing, and hybrid. Examples of each are shown in Figure 2.
- A fixed wing sUAS generates lift using the vehicle’s forward airspeed and the upward force caused by the shape of its wings. Normally, one or more propellers provide thrust for forward motion. Motion is in the direction of the nose of the sUAS and is needed to maintain lift; thus, fixed-wing sUASs cannot hover.
11 This is the theoretical maximum height where the air becomes too thin for a rotary-wing sUAS to continue flying. Practically, limitations on battery power and high-altitude winds would prevent a hobby sUAS from reaching this height.
- A rotary wing sUAS is a multi-rotor helicopter that is lifted and propelled by its rotors (vertically oriented propellers). Its lift is generated by a set of rotors. Horizontal motion is caused by tilting the rotors, and it can be in any direction. Rotary-wing sUASs normally have four to eight rotors. Like a helicopter, a rotary-wing sUAS can hover.
- A hybrid sUAS may be a vertical take-off and landing sUAS. For example, it may take off like a rotary-wing sUAS and then turn its entire airframe to fly like a fixed-wing sUAS. It may also be a combination of more diverse mobility functions like one that transitions back and forth between a sUAS and a small unmanned underwater (or surface) vehicle, or between a sUAS and a small, unmanned ground vehicle.
The operating modes for sUASs vary widely, depending on the maturity of autonomous behavior of a given sUAS and the needs and abilities of the human operator(s). The modes, discussed below, range from single, line-of-sight (LOS), remote-controlled sUASs to collaborative groups and swarms of fully autonomous sUASs.
Single sUAS with Varying Levels of Autonomy
Single sUAS with varying levels of autonomy include the following:
- Wired or wireless, LOS, remote control.12An operator controls all operations of the sUAS. The same operator or an observer near the operator must have a clear LOS with the sUAS to understand its three-dimensional location and orientation, especially with respect to its immediate surroundings, while controlling the movements of the sUAS.
- Wireless, non-LOS, remote control. An operator controls all operations of the sUAS. However, a direct LOS is not necessary to understand its three-dimensional location and orientation, especially with respect to its immediate surroundings, while controlling its
12Remote control refers to a device that is fully operated and controlled by a human using a tethered, untethered, or radio frequency/electro optical link.
movements. Onboard sensors generate digital information (e.g., video, text, and graphical information) to enable the operator to understand the location of the sUAS with respect to its surroundings and the operator while the operator controls the movement of the sUAS.
- Semi-autonomous. The sUAS can perform very limited control activities to enhance the ability of the operator to perform other tasks. For example, it may automatically go into a hover when the operator stops inputting commands to observe video from the sUAS, or the sUAS may automatically avoid obstacles while being flown by the operator. However, the sUAS can perform very few tasks on its own without accompanying operator input. A communications link is often used by the operator.
- Nearly full autonomy. The sUAS can perform many automated tasks, such as automatic flight control (including obstacle avoidance), engine control (for complex flight dynamics and hovering), target recognition, and target tracking (e.g., DJI’s Active Track function). However, the automated tasks are still activated or deactivated by the operator, and, if activated, will function without the specific knowledge of or control by the operator. An operator may direct the actions of individual or multiple sUASs in a supervisory role, especially in the execution of missions.
- Fully autonomous. Individual or large numbers of sUASs that require no human intervention to perform tasks, especially complex tasks such as planning and executing missions, navigating without GPS, avoiding obstacles, etc. The operator will assign missions, occasionally supervise the execution of missions, and be part of a manned-unmanned team.
Operator-Enabled, Coordinated sUASs
The concept of operations (CONOPS) for operator-enabled, coordinated sUASs, is that two or more operators of single sUASs coordinate their efforts before and/or during a mission to accomplish mission tasks. This CONOPS will most likely use remote-controlled or semi-autonomous sUASs.
Software-Enabled, Coordinated sUASs
The software-enabled, coordinated sUAS CONOPS is for one operator to control two or more semi- and/or nearly fully autonomous sUASs to accomplish a mission. However, these sUASs will operate independently of each other, and the operator will control each of them individually either by pre-programming or by controlling them during flight. Pre-programming (e.g., when launched, each flies to its own specified, pre-programmed location) may enable a single operator to coordinate the flight paths of numerous (more than 40) sUASs. If a change in mission requires dynamic reprogramming or coordination during flight, this will significantly task the cognitive abilities of the operator, thus reducing the number of sUASs being controlled for a particular mission from more than 40 to perhaps as few as 5, depending on the skills of the operator. This CONOPS will most likely use wireless, non-LOS, remote controlled, semi-autonomous, or nearly fully autonomous sUASs.
Swarm of sUASs
A swarm is a larger number (40, but maybe hundreds) of sUASs all following the same simple rules to achieve a goal (Aoki, 1982; Huth and Wissel, 1992; Reynolds, 1987). As the number of individual sUASs increases in a single swarm, humans lose the ability to track individual sUASs and begin to perceive multiple sUASs as a single entity (Seiffert et al., 2015). While it is not entirely clear at what number of entities this perceptual transition occurs, it is believed that the tipping point is about 40 sUASs. In current experimental swarms, an operator can control 40 to 100 sUASs. In the future, an
The key to a swarm is that the entire group appears to act as a single unit, but the individual sUASs actually act as distributed, local controllers (Aoki, 1982; Reynolds, 1987; Barca and Sekercioglu, 2013). The individual sUASs behave like a collective organism, sharing one distributed brain for decision-making and adapting to each other, like swarms in nature (e.g., flocks of birds and schools of fish). Each sUAS uses its software-based intelligence to coordinate its location among other sUASs and execute its localized behavior. The sUASs need not all be of the same type.
Not all members of the swarm will know the assigned mission. There may be leaders in a swarm with knowledge of the mission. The remainder of the swarm executes their standard localized behaviors without knowing the mission objectives, but as an entity the swarm executes the mission. For example, an operator may direct swarm leaders (Tiwari et al., 2017; Kolpas et al., 2013) to “search this area,” and the swarm automatically coordinates its individual members to accomplish the mission.
Communications among swarm members can be explicit (e.g., the exchange of trivial messages, like “I found the target”), but they will mostly be implicit (e.g., using onboard sensors to determine position relative to other nearby sUASs) (Haque et al., 2016). The models of communications from the biological swarm literature, mostly passive-based sensing, are metric (communicate with any sUAS within a specified distance) (Aoki, 1982), topological (based on number of aircraft, not on distance) (Ballerini et al., 2008), or visual/perceptual (react to all sUAS within LOS) (Standburg-Peshkin et al., 2013).
While swarming, each sUAS will position itself relative to other sUASs, establishing minimum relative distances between sUASs. These distances may be created by a combination of repulsion (moving away from nearby sUASs), orientation (determining one’s location relative to other sUASs), or attraction (moving toward nearby sUASs) (Aoki, 1982; Huth and Wissel, 1992; Reynolds, 1987). These sUASs will also change their location to improve their passive sensing outcomes. This swarm CONOPS would use nearly fully autonomous sUASs.
Collaborative Groups (<40) and Collaborative Swarms (≥40) of sUASs
In the CONOPS for collaborative groups (<40) and collaborative swarms (≥40) of sUASs, multiple sUASs (either all the same or different types) can perform sophisticated tasks as part of a team through data sharing, communications, and synchronization of actions, and even dynamically reassigning missions (e.g., sUAS-target pairings) to take advantage of the capabilities and physical location of team members and abide by established rules of engagement. Collaboration is accomplished through distributed, platform-based interactions—that is, individual platforms can contribute to the overall decision.
As an example, a group or swarm of collaborative sUASs approach a target area, and that approach may be from different directions and at different altitudes. Based on the priority of identified targets and the lethal capability of each sUAS, the group or swarm members will determine which sUASs attack particular targets (e.g., attacking high-priority targets with more than one sUAS).
A collaborative group or swarm will have a human commander who assigns its mission and, for parts of the mission, may release the sUASs to their fully autonomous mode in which they function with no human intervention. More advanced collaborative groups or swarms of sUASs may solicit nearby friendly ground and air platforms to join the group to accomplish an assigned mission or to adapt to dynamic situations, such as unanticipated threats. Additionally, as these advanced collaborative groups or swarms are reduced by attrition, they may consolidate forces to create new collaborative groups or swarms. This CONOPS will use fully autonomous sUASs.
The Army’s definitions for coordinated, swarm, and collaborative groups or swarms are significantly different from, and much less detailed than, the committee’s definitions. For CONOPS involving multiple sUASs, the Army uses two terms: saturation and swarm. For the Army, saturation is the use of a small group of operator-enabled or software-enabled coordinated sUASs to attack a target. The thought is that multiple sUASs will “saturate” the target and, hopefully, overcome counter-sUAS systems.
The Army defines a swarm as follows:
Swarming is a method of operations where large numbers of autonomous systems actively coordinate their actions to achieve operational outcomes. Swarming overwhelms targets by using mass and attrition in combination with decentralized maneuvers or combined fires from multiple directions (MASD ARCIC, 2017).
The committee does not view swarming as a method but rather as a cooperative behavior capability for a large number of semi-autonomous sUASs. When cooperative behavior is enhanced, utilizing fully autonomous sUASs, collaborative groups and swarms become possible. If the Army focuses its counter-sUAS efforts on its definitions of saturation and swarm, it may miss technical opportunities to develop overmatching counter-swarm capabilities.
sUASs generally fall into three categories of customization: consumer (i.e., non-customized, ready to fly), modified consumer (i.e., some level of customization), and customized (i.e., built from scratch).
A consumer sUAS (i.e., non-customized, ready to fly) is one that can be purchased in a store or online already assembled. The user has little to no understanding of the software and hardware operation and integration or how the components (parts of the sUAS) interact. Operations are limited to the advertised capabilities of the commercial sUAS and are limited by published controls (e.g., geo-fencing to prevent flying near sensitive areas such as an airport). These are generally of low to moderate cost ($1,000). An example is ISIS buying a DJI Phantom sUAS and using its integral camera for conducting surveillance missions.
A modified consumer sUAS (i.e., some level of customization) is assembled using sUAS components available in a store or online. The user integrates the components together, similar to building a model. This type of customization is limited by the commercial availability of sUAS components. The user has some understanding of component interactions and the science and engineering behind component functions. Modified consumer sUASs may exceed the limitations of commercial capabilities and published controls. These are generally of moderate to moderately high cost ($10,000). An example is ISIS’s “drone factories” where mostly fixed-wing sUASs were assembled using commercially available components.
A customized sUAS has some or all of its components designed, built, and tested by or for a user. This approach is not limited to the commercial availability of sUAS components for a modified consumer sUAS. This sUAS category requires an in-depth understanding of multiple technical fields. Customized sUASs can conduct highly sophisticated operations. These sUASs are generally of moderate to very high cost ($100,000). Threat examples are limited but include a report of wing modifications to enhance flight performance and many reports of customizing payloads for dropping grenades and small bombs. The committee is aware of very successful customization efforts by university students (IMechE, 2017; UAV Challenge, 2017; AUVSI SUAS, 2017). This level of customization is still short of the capabilities of nation states.
Countering sUAS(s) is the use of counter-sUAS materiel systems; tactics, techniques, and procedures (TTPs); and other approaches to prevent a sUAS from accomplishing its mission.
- A counter-sUAS materiel system is a dedicated, physical counter-sUAS system(s) used to implement the following kill chain;
- Detect, locate, and track potential targets;
- Identify, classify, and evaluate targets as sUASs;
- Engage and defeat (neutralize) sUASs;
- Verify the response through battle damage assessment; and
- Clean up and recovery.13
- Active TTPs involve the reemphasis of existing mission tasks (e.g., scanning the horizon), improving firing techniques with organic weapons (e.g., rifles) or modified organic weapons (e.g., a multi-purpose 40 mm round), or other actions to defeat sUASs.
- Passive TTPs involve the use of camouflage, stealth, decoys, or other actions to prevent the sUAS from locating the friendly force.
Other counter-sUAS approaches that consider doctrine, organization, training, leadership, personnel, facilities, and policy (i.e. DOT_LPF-P),14 for example,
- Doctrine (e.g., use of distributed operations),
- Organization (e.g., integration of air defense units in rifle companies),
- Training (e.g., including the equivalent of skeet training as part of basic training),
- Leadership (e.g., training in the integration of counter-sUAS considerations into operations),
- Personnel skills (e.g., establishing career fields in counter-sUAS),
- Facilities (e.g., hardening forward operating bases against sUAs), and
- Policy (e.g., balancing realism in training with safety).
The performance of a materiel counter-sUAS system will depend on a large number of factors. These include the following:
- Target acquisition capability and range;
- Engagement range;
- Effect on target(s);
- Performance characteristics of threat sUASs (including autonomy);
- Level of sUAS customization;
- Vulnerability of the components of sUASs (as identified in the taxonomy in Appendix D);
- Governing weapons control order;
- Threat TTPs (including the use of swarms and collaborative groups);
- The operational environment (e.g., open versus urban terrain);
- Lighting conditions; and
- If the counter-sUAS system is not automated,
- Operator skills,
13 This bullet was added by the committee as an additional kill chain activity in order to include functions the committee believes are important to the counter-sUAS fight.
14 Materiel is omitted from what is the usual Doctrine, Organization, Training, Materiel, Leadership, Personnel, Facilities, and Policy (i.e., DOTMLPF-P) formulation because it is addressed in the first bullet as “counter-sUAS materiel system.”
- Training, and
- Physical and mental performance.
Additional terms are defined in Appendix C. These terms include the following:
- Deep magazine versus stowed kills,
- Target acquisition levels,
- Neutralization, and
- Weapons control order.
The committee conducted data gathering at all five of its meetings, with the fourth meeting focused on gathering restricted information. Committee activities and the organizations they heard from are listed in Appendix A. In addition to approaching various Army organizations to learn about their view of the sUAS threat and current and planned counter-sUAS efforts, the committee also reached out to the U.S. Navy, the USMC, the Joint Improvised-Threat Defeat Organization, and to DHS. Of particular note was an opportunity for the committee to meet with 12 officers and enlisted members of the 3rd Battalion, 5th Marines, to discuss their experiences and lessons learned from their counter-sUAS experiments held in the training areas at 29 Palms Marine Corps Base. These Marines provided insightful information, providing the committee a very thought-provoking user perspective.
Data gathering for this report ended with the fifth meeting and was cut off around April 28, 2017. After that point in time, the development of sUASs and counter-sUAS systems continued to advance at a very rapid pace. There has been a significant amount of discussion and activity related to the adversarial use of sUASs and the development of counter-sUAS systems since April. Due to the necessity to stop data gathering so the report could be finalized and reviewed, most information since April is not included in this report. Small amounts of data were collected to clarify the discussion, findings, and recommendations developed in April.
The restricted report is organized into six chapters and five supporting appendixes. The chapters and appendixes provide the following information:
- Chapter 1 includes the statement of task, origins of the study, timeframe taxonomy, modes of sUAS utilization (including descriptions of coordinated groups, swarms, and collaborative swarms or groups), levels of customization,’ definitions, and a functional decomposition of sUASs, with additional definitions and discussions in Appendixes C and D.
- Chapter 2 includes a discussion of performance capabilities of sUASs, including a committee assessment of likely improvements in the immediate, imminent, and emerging timeframes. It also includes a high-level roadmap of those capabilities.
- Chapter 3 discusses human performance considerations, as well as logistics and operational usability/utility considerations.
- Chapter 4 discusses current counter-sUAS efforts and assesses counter-sUAS approaches to identify counter-sUAS capability needs. This chapter also includes a quality function deployment (QFD) analysis, with accompanying QFD details in Appendix E.
- Chapter 5 focuses primarily on recommending science and technology development efforts, but also discusses non-materiel approaches in areas of doctrine, organization, training, leadership, personnel, facilities, and policy.
- Chapter 6 discusses the DHS concern about sUAS activities in the homeland and potential areas of interest from Chapter 5.
- Appendix A discusses committee activities.
- Appendix B presents the committee member biographies.
- Appendix C presents some definitions used by the committee that were omitted from Chapter 1 for the sake of brevity.
- Appendix D presents a decomposition of sUAS capabilities.
- Appendix E presents a detailed discussion of the rating factors supporting the QFD tables in Chapter 4. ·
Many findings and recommendations are identified and detailed in the restricted report, including seven high-impact recommendations.
60 Minutes. 2017. Capturing the swarm. CBS News. http://www.cbsnews.com/news/60-minutes-capturing-the-perdix-drone-swarm/.
Aoki, I. 1982. A simulation study on the schooling mechanism in fish. Nihon-suisan-gakkai-shi 48(8):1081-1088.
Atherton, K.D. 2016. IED drone kills Kurdish soldiers, French commandos. Popular Science.https://www.popsci.com/booby-trapped-isis-drone-kills-kurdish-soldiers-french-commandos.
AUVSI SUAS (Association for Unmanned Vehicle Systems International Student Unmanned Aerial Systems). 2017. Student Unmanned Aerial Systems Challenge. http://www.auvsi-suas.org/competitions/2017/.
Ballerini, M., N. Cabibbo, R. Candelier, A. Cavagna, E. Cisbani, I. Giardina, V. Lecomte, A. Orlandi, G. Parisi, A. Procaccini, M. Viale, and V. Zdravkovic. 2008. Interaction ruling animal collective behavior depends on topological rather than metric distance: Evidence from a field study. Proceedings of the National Academy of Sciences US.A. 105(4):1232-1237.
Barca, J.C., and Y.A. Sekercioglu. 2013. Swarm robotics reviewed. Robotica 31(3):345-359.
Barrett, B. 2016. When Good Drones Go Bad. Wired.com. January 18. https://www.wired.com/2016/01/when-good-drones-go-bad/.
Bedard, P. 2015. Drone sales surge 167% to 4.3 million, U.S. leads but China catching up. Washington Examiner. May 29. http://www.washingtonexaminer.com/drone-sales-surge-167-to-43-million-us-leads-but-china-catching-up/article/2565240.
Brown, D.S., M.A. Goodrich, S.-Y. Jung, and S. Kerman. 2016. Two invariants of human-swarm interaction. Journal of Human-Robot Interaction 5(1):1-31.
Copp, T. 2016. Pentagon worried about drone threat to nuclear sites. Stars and Stripes. September 27. https://www.stripes.com/news/pentagon-worried-about-drone-threat-to-nuclear-sites-1.431196.
DoD (Department of Defense). 2011. Unmanned Aircraft System Airspace Integration Plan, version 2.0. Arlington County, Va.: UAS Task Force, Airspace Integration Integrated Product Team.
Freedberg, Jr., S.J. 2015. Russian drone threat: Army seeks Ukraine lessons. Breaking Defense. October 14. https://breakingdefense.com/2015/10/russian-drone-threat-army-seeks-ukraine-lessons/.
French, S. 2016. Drone sales in the U.S. more than doubled in the past year. MarketWatch. May 28. https://www.marketwatch.com/story/drone-sales-in-the-us-more-than-doubled-in-the-past-year-2016-05-27.
Gillis, J. 2017. In over their heads: U.S. ground forces are dangerously unprepared for enemy drones. War
Goodrich, M.A., B. Pendleton, S. Kerman, and P. Sujit. 2012. What Types of Interactions Do Bio-Inspired Robot Swarms and Flocks Afford a Human? Pp. 105-112 in Proceedings of Robotics, Science and Systems. Sydney, Australia: Robotics: Science, and Systems.
Hambling, D. 2016. If drone swarms are the future, China may be winning. Popular Mechanics. December 23. https://www.popularmechanics.com/military/research/a24494/chinese-drones-swarms/.
Haque, M., C. Ren, E. Baker, D. Kirkpatrick, and J.A. Adams. 2016. Analysis of Swarm Communication Models. Pp. 29-37 in Proceedings of the 6th International Workshop on Combo of Intelligent Methods and Applications. https://www.researchgate.net/profile/Zhidong_Deng/publication/308946227_Generic_Convolutional_Neural_Network_with_Random_Pooling_Area/links/57f919eb08ae91deaa6166d5.pdf#page=39.
Huth, A., and C. Wissel. 1992. The simulation of the movement of fish schools. Journal of Theoretical Biology 156(3):365-385.
IMechE (Institution of Mechanical Engineers). 2016. The Unmanned Aircraft Systems Challenge. https://www.imeche.org/get-involved/young-members-network/auasc.
Karber, P. 2015. “Lessons Learned” from the Russo-Ukrainian War: Personal Observations. July 8. Draft. The Potomac Foundation. https://prodev2go.files.wordpress.com/2015/10/rus-ukr-lessons-draft.pdf.
Kolling, A., K. Sycara, S. Nunnally, and M. Lewis. 2013. Human swarm interaction: An experimental study of two types of interaction with foraging swarms. Journal of Human-Robot Interaction 2(2):103-128.
Kolling, A., P. Walker, N. Chakraborty, K. Sycara, and M. Lewis. 2016. Human interaction with robot swarms: A survey. IEEE Transactions on Human-Machine Systems 46(1):9-26.
Kolpas, A., M. Busch, H. Li, I.D. Couzin, L. Petzold, and J. Moehlis. 2013. How the spatial position of individuals affects their influence on swarms: A numerical comparison of two popular swarm dynamics models. PLoS ONE 8(3):e58525.
MASD ARCIC (Maneuver, Aviation, and Soldier Division, Army Capabilities Integration Center). 2017. The U.S. Army Robotic and Autonomous Systems Strategy. http://www.arcic.army.mil/App_Documents/RAS_Strategy.pdf.
Murphy, M. 2016. UK is crowdsourcing its swarming drone attacks. Defense One. October 4. http://www.defenseone.com/technology/2016/10/uk-crowdsourcing-its-swarming-drone-attacks/132082/.
Remote Control Project. 2016. Hostile Drones: The Hostile Use of Drones by Non-State Actors Against British Targets. U.K., London: Oxford Research Group. January. http://remotecontrolproject.org/wp-content/uploads/2016/01/Hostile-use-of-drones-report_open-briefing.pdf.
Reynolds, C.W. 1987. Flocks, herds, and schools: A distributed behavioral model. Computer Graphics 21(4):25-34.
Seiffert, A.E., S.T. Hayes, C.E. Harriott, and J.A. Adams. Motion Perception of Biological Swarms. Pp. 2128-2133 in Proceedings of the Annual Cognitive Science Society Meeting. Pasadena, Calif.: Cognitive Science Society.
Strandburg-Peshkin, A., C.R. Twomey, N.W.F. Bode, AB. Kao, Y. Katz, C.C. Ioannou, S.B. Rosenthal, C.J. Tomey, H.S. Wu, and S.A. Levin. 2013. Visual sensory networks and effective information transfer in animal groups. Current Biology 23(17):R709-R711.
Tiwari, R., P. Jain, S. Butail, P.B. Sujit, and M.A. Goodrich. 2017. Effect of Leader Placement on Robotic Swarm Control. In Proceedings of the 16th International Conference on Autonomous Agents and Multiagent Systems. Sao Paulo, Brazil: International Foundation for Autonomous Agents and Multiagent Systems.
Tucker, P. 2015. Did NATO forces just down a Russian drone over Turkey? Defense One.http://www.defenseone.com/threats/2015/10/did-nato-forces-just-down-russian-drone-over-turkey/122882/.
UAV Challenge. 2017. Announcing the 2017 UAV Challenge Airborne Delivery Challenge.
Walker, P., S.A. Amraii, N. Chakraborty, M. Lewis, and K. Sycara. 2014. Human Control of Robot Swarms with Dynamic Leaders. Pp. 1108-1113 in Proceedings of 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Piscataway, N.J.: IEEE.
Watson, B. 2017. The drones of ISIS. Defense One.http://www.defenseone.com/technology/2017/01/drones-isis/134542/.
This page intentionally left blank.
This page intentionally left blank.