Context

Australia’s strategic geography will continue to present new challenges and opportunities into the future. The nation will remain an enormous landmass with a small population at the bottom of Asia. Yet, no longer is our region a strategic backwater with geographic depth. Our historic geographic advantage is being eroded by grey zone actions, the reach and lethality of modern weapons, the potency of anti-access and area-denial (A2AD) systems, and the broader modernisation of other nations’ military forces. These considerations suggest a military problem: how can the vast maritime expanses, northern archipelagic region, and congested littorals afford Australia offset and strategic depth in the future?

Australia’s historical economic and technological advantages will likely have eroded by 2040. Australia’s economy is being dwarfed by more populous regional nations such as Indonesia; Australian GDP is predicted to be one third or less of Indonesia’s GDP by 2030.[1] Further, the proliferation of technology coupled with the ever-increasing cost and acquisition time to field small numbers of so-called ‘exquisite platforms’ is eroding Australia’s strategic edge. All the while, the trends postulated in the concept of Accelerated Warfare will continue to shape the future operating environment. Demographic reality dictates that Australia will continue to field a numerically modest military. This suggests our second challenge: how can Australia offset these conditions to generate mass when economically and technologically surpassed within the region?

Swarming

The large mass of autonomous systems interoperating collectively to act and respond in a coordinated effort to provide an overwhelming effect.

Australian Defence Glossary

For the purpose of this synopsis, swarming is defined as the application of swarm intelligence research to swarm robotics. Adding some specificity to the definition in the Australian Defence Glossary, a robot swarm is 1) three or more robots, 2) which have limited or no human control, and 3) which perform tasks cooperatively.[2] This paper transcends the common conceptualisation of swarming ‘drones’, or Unmanned Aerial Systems, and instead discusses Uninhabited Vehicles (UxV) operating in and across all warfighting domains (land, air, maritime, space and cyber). A homogeneous swarm is three or more UxVs of the same type undertaking the same task in cooperation. A heterogeneous swarm contains independent UxVs delivering alternate effects in unison. UxVs can also be deployed in small manoeuvre units, called ‘clusters’, which can surge to swarm and provide effects in time and space.

‘Exquisite platforms’ incorporate a host of capabilities within a single system. It is probable that a technological edge based predominantly on exquisite platforms is unlikely to be a sustainable approach for Australia in 2040 and beyond. As an alternative, UxVs produced for heterogeneous swarming can replicate the effects of exquisite platforms by fielding numerous types of small, independent platforms which may individually have a single purpose (ISR, EW, air-defence, anti-tank etc.), but are able to mass to provide collective effects. “Small, smart and cheap”[3] UxVs could be produced in the hundreds or thousands, both before and during conflict, to overwhelm an adversary, generate advantage, and sustain a war effort. This idea resonates with the concepts for UxV’s being developed by partner militaries, as emblemised by the U.S. Defense Advanced Research Projects Agency’s “Mosaic Warfare” concept.[4]

The technical challenge of achieving swarm Artificial Intelligence (AI) within swarm robotics is extremely high. However, the race is on with numerous nations pursuing military employability of swarm technology.

Supporting Unified Domain Operations 2040: Small, Smart, Cheap and Many

Army’s future operating concept, Unified Domain Operations 2040 (UDO40), recognises the requirement to conceptualise a single warfighting domain wherein Australia must realise asymmetric advantage from its people, weapons and systems.UDO40 seeks to deter or defeat peer (plus) threat attacks on Australia or its interests by threatening to, or actually imposing, costs that exceed any perceived advantage. In this vein, heterogeneous swarming (which potentially includes aerial, ground, surface, and/or underwater UxV clusters able to surge and coordinate effects) provides an example of Unified Domain Operations in action. Further, swarming could complement Australia’s geography and strategic approaches by using swarm ‘mass’ to delay an adversary or hinder their projection of force into the region.

Swarming offers Australia a potential way to maintain its offset strategy[5] and continue to generate advantage. Conceptually, UxV swarming could provide a cheap, expendable asymmetric approach to detecting, deterring, denying and/or defeating adversaries within Australia’s near region. The following indicative list describes the tactical and operational utility of swarming to the ADF, conceptualised through the six combat functions.

Know. A key advantage offered by swarming is the ability to field large numbers of systems with limited human operator intervention, increasing the operational effect while reducing the required human capital. Such swarms can cover swathes of territory while internally coordinating their own efforts. This could include saturating Australia’s northern approaches or maritime choke points with aerial, surface or underwater UxVs to detect, track and if required disrupt or defeat adversary submarines or vessels. Aerial, ground and surface UxVs could be employed to seek out enemy high value targets, even within a contested A2AD environment, to then queue strike effects. Massed UxVs could operate with agility in congested terrain such as in jungle or megacities by swarming with hundreds or thousands of small aerial and ground ISR UxVs.

Adapt. Swarming can provide support to C2 as a result of the know function, thus enhancing commanders’ adaptability. The standoff capability created by deploying ISR swarms into Australia’s northern approaches might negate the Joint Force requirement to deploy into the close battlespace until greater situational awareness has been achieved. This could afford a greater level of adaptability in the early stages of a conflict. The ability of swarming systems to disaggregate as clusters or individual UxVs, and then re-aggregate can create an adaptability and tempo that is hard to counter. When swarm robotics can achieve not only semi-autonomous but also autonomous swarming (that is human-in, human-on, or human-out-of the loop), UxVs could operate and adapt from C2 nodes within a swarm, even when in a communications degraded or denied environment. This would directly support Hierarchical Command – Agile Control which is the central idea of Australia’s future command and control concept.[6] Such adaptability could enable a swarm to continue to seek out enemy systems or platforms that might be contributing to an A2AD effect, or degrading friendly communications, and reduce or remove these adversary systems.

Shape. Swarming UxV’s can be a decoy. Massed UxVs can confuse physical and electronic methods of detection. Swarms of ground UxVs could deceive an adversary into thinking a larger vehicle formation is moving through an area, or swarmed aerial UxVs could present the same electronic cross-section of a helicopter to confuse adversary sensors. UxVs enabled with EW deception capabilities could shape adversary forces away from the friendly main effort by presenting surges in emissions across the battlespace. These UxVs could also disaggregate into clusters to present overwhelming signals traffic to overload an adversary’s sensors, masking true friendly signals, or to shape the adversary into thinking they are facing a larger force. UxVs could also be employed kinetically to shape an adversary’s course of action. Finally, the mere presence or threat of swarming could alter an adversary’s calculus.

Strike. Swarming makes kinetic and non-kinetic contributions to strike effects. At its most basic are swarming munitions, or ‘swarming by fire’, whereby loitering autonomous munitions can find a target and then swarm to destroy it. However, the maximum utility of swarming is more likely to be in autonomous UxVs operating in conjunction with other systems in a heterogeneous swarm, or ‘swarming by force’. An AI enabled swarm could produce targeting solutions from ISR UxVs to queue UxV strike platforms or other standoff weapons via ‘best sensor, best effector’ methodology. This could be a strike from within the swarm by kinetic UxVs, or non-kinetic UxVs delivering jamming or other EW or cyber effects. A conceptual ‘kill mesh’ could be achieved with hundreds or thousands of these UxVs projecting into a contested A2AD environment in Australia’s near region to deliver strike effects. This could create standoff, asymmetrically negate adversary capabilities with expendable friendly platforms, leverage Australia’s strategic geography, and thus protect the force. Furthermore, by achieving mass with UxVs, the numbers of humans required in the close and deep battlespace could be reduced, alleviating the corresponding logistics burden.

Shield. Swarming can create standoff to protect the joint force. Future ‘advance to contact’ could be led by a vanguard of swarming UxVs to absorb the initial contact. Swarms could also be employed to defensively ‘mine’ maritime ports and choke points, as well as airports or the air domain above friendly installations or facilities. For example, a swarm of aerial and ground UxVs could remain within or adjacent to a friendly base in a hibernation mode to provide a shield effect. Upon sensing the approach of adversary aircraft, the aerial systems could launch an aerial minefield with a reactive swarm of small obstacles or explosive devises. The ground UxVs could track and direct the aerial swarm and/or launch anti-air munitions in unison. Further, our own swarms can counter that of the adversary by interdicting massed threats. .[7]

Sustain. Resupply could be partially facilitated by swarms of UxVs that are expendable and difficult to intercept, or at least challenging to destroy in their entirety. If the UxVs developed for swarming were kept small, cheap and produced nationally, the ability to sustain or replace swarming formations could be extensive. Once the UxVs were operating in the battlespace, the predominant considerations would be munitions and power requirements. Smaller systems could be projected into the battlespace by larger aerial, surface or underwater ‘mothership’ UxVs with larger power supplies. Power resupply could then be facilitated by UxVs returning to rear echelons, linking up with solar powered ‘recharge’ UxVs within a swarm, or returning to the larger UxV ‘mothership’ platforms providing a resupply node. Finally, the very concept of swarming lends itself to combat sustainability as there is inherent resilience afforded by overwhelming an adversary with mass where the destruction of a central control node is not possible.

While beyond the scope of this synopsis, the concept of swarming in support of an Australian offset strategy is important to explore. This exploration can be undertaken alongside Army’s partners. The concept would require inclusion in future sovereign industry capability policy as the technology on offer is well within Australia’s industrial capacity to grasp. Supported by a readily available source of UxV’s, many of the Army’s challenges in ‘scaling’ its capacity could be overcome. Finally, a concept is necessary such that the implications for AI are appropriately considered. Of relevance, the opportunity presented by swarming has been noted in the Army’s Robotic and Autonomous Systems Strategy and the ADF’s Joint Concept for Robotic and Autonomous Systems.[8]

While swarming will not be devoid of limitations and vulnerabilities, the potential asymmetric advantage offered by maturation and adoption into service of such a capability would be profound. It would also be timely given that numerous other nations are already pursuing this technology. Australia’s unique geostrategic circumstance and predicted future operating environment mean that swarming can significantly contribute to sustaining a capable offset strategy beyond 2040.

[1] Lowy Institute, Asia Power Index 2020, Canberra, 2020, https://power.lowyinstitute.org/data/future-resources/economic-size-203… (accessed 05 March 2021)

[2] R. Arnold et al, ‘What is a Robot Swarm: A Definition for Swarming Robotics,’ 2019 IEEE 10th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference, New York, 2019, p. 75

[3] T.X. Hammes, ‘Technologies Converge and Power Diffuses: The Evolution of Small, Smart, and Cheap Weapons’, Policy Analysis, Cato Institute, No. 786, 2016, https://cato.org/publications/policy-analysis/technologies-converge-pow… (accessed on 01 March 2021)

[4] Defense Advanced Research Projects Agency (DARPA), DARPA Tiles Together a Vision of Mosaic Warfare: Banking on Cost-effective Complexity to Overwhelm Adversaries, 2018, https://www.darpa.mil/work-with-us/darpa-tiles-together-a-vision-of-mos…, (accessed 11 March 2021); B. Jensen & J. Paschkewitz, ‘Mosaic Warfare: Small and Scalable are Beautiful’, War on the Rocks, 2019, https://warontherocks.com/2019/12/mosaic-warfare-small-scalable-are-bea…, (accessed 11 March 2021)

[5] I. Langford, ‘Australia’s Offset and A2/AD Strategies’, Parameters 47(1) Spring 2017, 2017

[6] Force Exploration Branch, ADF Concept for Command and Control of the Future Force (version 1.0), Department of Defence, Canberra, 2019

[7] Counter-swarming is conceptually as broad and complex as this discussion on Australia’s employment of swarming, but is beyond the scope of this paper.

[8] Future Land Warfare, Robotic and Autonomous Systems Strategy, Department of Defence, Canberra, 2018, p. 10; Force Exploration Branch, Concept for Robotic and Autonomous Systems (version 1.0), Department of Defence, Canberra, 2020, p. 38-39.

Introduction