Lynne Rienner Publishers Inc, Boulder CO, 2024, 257 pp
Hardcover ISBN: 9781685859817
Authors: Ash Rossiter and Peter Layton
Reviewed by: Callum Hamilton
Warfare in the Robotics Age examines the evolution of military robotics, how defence forces are adapting to increased autonomy in warfare, and the consequences this trend may have for international relations, statecraft, and deterrence. It advances the literature by linking the emergence of military robotic and autonomous systems (RAS) to Fourth Industrial Revolution (4IR) innovations. The authors’ research provides practical insights that will help RAS practitioners, warfighters and national security policymakers alike to navigate challenges for militaries as they seek to employ RAS.
Given the hype that can surround the analysis of RAS, the authors make a clear point of balancing opportunities for capability advantage with discussion of the challenges and uncertainty that remain in the field. To this effect, Warfare in the Robotics Age begins with the history of robotics in warfare, which it traces back to the use of wire-guided, remote-controlled ‘land torpedos’ in World War I. Despite the surprisingly long history of military robotics, the authors note several challenges which have slowed development. Until recently, robotics has struggled to find support within military services owing to a limited track record of success and an inability to translate technological development into manpower efficiencies.
Scepticism towards robotics is slowly fading as 4IR technologies build momentum in both civilian and military contexts, driven by interrelated advancements in artificial intelligence (AI), big data, cloud computing, and the internet of things. These technologies, as the authors describe, are bridging the physical and digital worlds by enabling robots to process unlabelled data, share information across systems, and self-optimise as they prompt action and change in the real world. Greater autonomy in robotics, or RAS, is disrupting military organisations, as the possible benefits for force scalability, survivability and efficiency come closer into view.
Today many forces around the world are experimenting with RAS—including Army’s Robotic and Autonomous Systems Implementation & Coordination Office (RICO)—and contributing to the growing body of evidence for its value to future forces. The experience of RICO agrees with the authors’ discussion:
Warfighters who can experiment with an operational prototype can more easily envisage the potential of the system than if the concept only remains intangible and theoretical … The prototypes provide opportunities to develop and refine new concepts of operation, … evolve operational requirements … and operate militarily useful quantities of prototype systems in realistic military demonstrations, and on that basis, to make an assessment of the military utility of the proposed capability.[1]
Further, experimentation activities provide an opportunity to resolve the uncertainties of RAS that would otherwise risk or obstruct long-term capability development programs.
Warfare in the Robotics Age introduces the concept of ‘prototype warfare’ to help make sense of the quickly evolving RAS ecosystem. Prototype warfare emphasises the rapid fielding of a small quantity of prototypes with soldier end users for assessment in militarily realistic experimentation activities. It provides a valuable opportunity to explore different ways of generating effects with technology before committing to a particular system, helps determine technical requirements, and reduces uncertainty around future capability. By approaching design and development as an evolving process rather than ‘one and done’, militaries might rapidly field a wide range of RAS platforms, placing additional resources behind those that provide an advantage in the strategic context of the day.
The authors contend that, in the future, prototype warfare may even enable continual innovation in military RAS, by extending the production process to the in-service phase. With warfighters at the centre of the production ecosystem, RAS platforms might be customised for specific missions or in response to changes at the operational and tactical levels. Importantly, RAS platforms that leverage AI and machine learning have a tendency for ‘concept drift’ as real-world data evolves over time in non-obvious ways. Performance monitoring and feedback, provided by soldier end users, will be key to enabling model refinement, retraining and domain adaptation to occur at the speed of relevance within operations. To seize the initiative in future conflicts, militaries will need the capability to continually adapt emerging technologies in disruptive ways and conduct assurance in significantly more compressed timeframes than has been necessary in the past.
While describing the potentially decisive advantages achievable by militaries that can continually innovate, Warfare in the Robotics Age always keeps sight of the challenges to RAS. In particular, it articulates the inherent tension between military bureaucracies (designed to provide uniformity) and the organisational agility needed to enable rapid prototyping and iteration. To explore competing RAS alternatives, militaries must accept enduring and substantial uncertainty. Not all concepts will realise the envisioned capability benefits. Some may only provide an advantage in certain operational and strategic contexts. Nevertheless, the authors argue that the potential capability and cost advantages of RAS create an imperative for states to accelerate concept development and avoid being on the wrong side of a capability asymmetry.
Warfare in the Robotics Age acknowledges that RAS may unlock mass and scalable effects for middle powers, such as Australia, that have previously not been able to compete against states with larger populations. Ukraine’s ability to rapidly deploy many small, cheap RAS platforms to attrit a larger Russian invading force lends weight to this line of thought. Although the book goes into great detail about developments in the United States and China, as undisputed global leaders in RAS, it would have been interesting to more deeply explore the asymmetric advantages that may fall to smaller states.
One topic the authors tread carefully around is the law and ethics of militarily employing RAS. This is not a criticism of Warfare in the Robotics Age. Much of the existing literature in this area is prone to generalisations and vagueness. While writers commonly note that law and ethics are clear challenges, it is rare to find practical solutions beyond the need to simply program legal and ethical decision-making into RAS. To the authors’ immense credit, Warfare in the Robotics Age proposes a different way forward based on the idea that only humans can ‘do’ responsibility and accountability. No level of human supervision will make a RAS concept acceptable if it cannot be reconciled with doctrine and responsible command. Only commanders (and operators) with an understanding of a RAS concept’s limitations will have the confidence needed to employ it across a range of tactical situations, noting their accountability for forces under their control.
The authors describe the challenge that integrating RAS in human-machine teams presents to developing testing and evaluation mechanisms, as well as tactics, training and procedures that build confidence in RAS concepts’ reliability. In particular, variable human interaction with RAS and AI-enabled machines will have a significant influence on the range of possible outputs that may need to be tested, evaluated and validated in training and operational environments. Answering how to build confidence in RAS, where unpredictability is a feature rather than a drawback, continues to be a key area for future research.
Warfare in the Robotics Age offers a compelling perspective on military robotics, with practical and easily digestible insights for RAS practitioners and subject matter enthusiasts alike. Crucially, Warfare in the Robotics Age addresses how and why military organisations (which are generally designed not to change) must embrace continual adaptation, and the uncertainty that comes with it, to realise the promise of RAS capability.
Endnotes
[1] Ash Rossiter and Peter Layton, Warfare in the Robotics Age (Boulder CO: Lynne Rienner Publishers Inc, 2024), p. 94.