Warfare in the Robotics Age

Rossiter and Layton bring their volume out with Lynne Rienner in 2025, into a defence-studies conversation that had shifted decisively in the years leading up to publication. The 2020 war in Nagorno-Karabakh, in which Azerbaijan used Turkish Bayraktar TB2s and Israeli loitering munitions to dismantle an Armenian ground force built around Soviet-era tanks and air defences, was the first widely absorbed shock. The full-scale Russian invasion of Ukraine two years later turned an intermittent argument about the future of unmanned systems into a running observation about the present. By the time the manuscript was finalised, tactical quadcopters were being manufactured in Ukrainian basements at a rate of millions per year, one-way attack drones from Iran were flying against European infrastructure through Russian hands, and Ukrainian uncrewed surface vessels had pushed the Russian Black Sea Fleet out of much of its home water. The authors write into that moment. Ash Rossiter is an associate professor at Khalifa University in Abu Dhabi, whose research focuses on security and emerging military technology in the Gulf; Peter Layton is a former Royal Australian Air Force officer now at the Griffith Asia Institute in Brisbane, with a long record of writing on air power, grand strategy, and the operational use of artificial intelligence. Their pairing lends the book a hybrid character — Rossiter’s regional grounding in the Middle East and Gulf drone industries alongside Layton’s air-power sensibility and interest in AI as an operational input rather than a research abstraction.

The central argument threaded through the volume is that robotics has crossed a threshold. It is no longer a supplement to the traditional force structures of major militaries but a defining feature of contemporary conflict, and the inherited framework of air, land, and sea power that has organised Western military thought since the mid-twentieth century now needs to be revised to accommodate machines that operate across domains and increasingly without continuous human direction. Rossiter and Layton do not argue that crewed platforms are obsolete or that autonomy has arrived at general intelligence. They argue something more modest and more consequential: that unmanned and semi-autonomous systems are now cheap enough, capable enough, and numerous enough to change the arithmetic of attrition, the geography of reach, and the industrial base a country needs to sustain a war. That claim carries implications for procurement, for doctrine, and for the political economy of defence. Small states with credible drone industries can hold larger neighbours at risk. Legacy platforms designed around a small number of exquisite, expensive units struggle against opponents that can generate loss faster than the platforms can be replaced. And the pace at which software-defined weapons iterate under combat conditions has begun to outrun the industrial cycles of Western defence procurement, which were built for aircraft and ships that spend more time on drawing boards than in service.

The book is organised as a structured tour of the robotic turn rather than a narrative of any single campaign. Aerial systems receive the most sustained treatment, tracked from the American Predator and Reaper lineage through the Israeli Heron and Harop, the Turkish Bayraktar TB2 and Anka, and out to the enormous population of small first-person-view racers and one-way attack munitions that now define the tactical edge of the Ukrainian front. Ground robotics, an area that has developed more slowly than the air, is treated across logistics mules, engineering and sapper platforms, remote weapons stations, and the armed unmanned ground vehicles that Russian and Ukrainian units have begun deploying for close-range assault. Maritime chapters cover uncrewed surface vessels of the Magura and Sea Baby type, the American Overlord unmanned vessel programme and its successors, and the deeper problem of undersea autonomy, where communications constraints and the difficulty of persistent power complicate the transfer of lessons from the air. The volume also gives sustained attention to artificial intelligence as a tool inside the kill chain rather than a science-fiction actor at the end of one — machine target recognition in imagery and radar, swarming and cooperative behaviours, human-machine teaming concepts of the kind embodied in the Australian MQ-28 Ghost Bat and the American CCA programme, and the software layer that turns a piece of moving hardware into a weapon. Doctrine, procurement reform, ethics and the legality of autonomous targeting, and the responses of Russia and China to Western unmanned advantage each get their own space. The scaffolding is that of a graduate seminar text: the reader is being walked around a field, not carried through a story.

The evidence is drawn from the wars of the last decade and the industries that have emerged around them. The Turkish drone industry is treated as one of the more consequential geopolitical developments of the 2020s, with Baykar’s Bayraktar TB2 and the newer TB3 and Kizilelma feeding a Turkish foreign policy that has used drone exports as a form of influence in Ukraine, Libya, Ethiopia, and Central Asia. Israel appears as the older and more diversified drone power — the country whose defence firms pioneered the tactical UAV in the 1970s and 1980s and whose loitering munitions, from Israel Aerospace Industries’ Harpy and Harop through UVision’s Hero series, established the category before it acquired the awkward Anglo-American name of one-way attack munition. The United Arab Emirates receives attention as an emerging exporter with its EDGE Group and Halcon subsidiaries producing indigenous designs, a development that Rossiter has covered in his own regional work. Iran is treated as the outlier that industrialised a cheap airframe — the Shahed-136 — and turned it into the first mass-produced strategic weapon of its class, licensed and produced in Russia as the Geran-2 for the sustained campaign against Ukrainian energy infrastructure. The Ukrainian side of that same war supplies most of the book’s tactical detail: FPV racers converted into loitering munitions with a handful of dollars of parts, night-flying bomber quadcopters dropping grenades and mines with terrifying reliability, the Magura V5 and Sea Baby surface vessels that struck the Ivanovets and the Sergey Kotov and forced the redeployment of the Black Sea Fleet from Sevastopol to Novorossiysk, and the emergence of a civil-military drone economy in which volunteer workshops produce hardware that Western defence primes cannot match on either price or delivery. On the Chinese side, the treatment covers the CH-4 and Wing Loong export lines, the more capable GJ-11 Sharp Sword flying wing, the People’s Liberation Army’s interest in swarming and unmanned aviation as a way of complicating American operations inside the first island chain, and the doctrinal writing coming out of the Academy of Military Sciences on intelligentised warfare.

Across these cases the authors return to a set of recurring problems. The first is cost. A Shahed-136 costs perhaps twenty to fifty thousand dollars to produce and is being shot down by air defence missiles that cost hundreds of thousands or millions each, an economics that makes any sustained defensive campaign untenable without cheaper interceptors. The second is signature. Small drones are hard to see on radar, hard to distinguish from birds when they are, and hard to engage at scale, which has renewed interest in electromagnetic and directed-energy defences and in low-cost kinetic interceptors of the kind the Ukrainians have improvised out of light aircraft and other drones. The third is the human place in the loop. The book treats the debate about autonomous targeting with more care than many popular accounts, drawing on the arguments running through the United Nations Group of Governmental Experts on Lethal Autonomous Weapons Systems and the divergent positions of major powers on whether and how to constrain machine decisions to kill. The fourth is industry. Legacy defence primes, organised around long production runs of exquisite platforms and cost-plus contracting, are not the firms that have produced the drones defining the current war; the new production comes from firms and workshops that iterate on the software side at commercial speeds, and the authors argue that Western militaries will have to build institutional relationships with a different kind of supplier if they want to sustain any of the capabilities they observe in Ukraine.

Reception has followed the arc typical of academic volumes that arrive at the leading edge of a running conversation. Reviewers have credited the book with the useful step of pulling the aerial, ground, and maritime strands into a single frame rather than leaving them as separate literatures, and with taking the industrial-base question seriously rather than treating it as a footnote to platform description. Where the pushback has come, it has come from two directions. Analysts closer to the front — the community around David Hambling’s writing, Sam Bendett’s tracking of Russian military robotics, or the reporting collected on the drone side of the Institute for the Study of War’s output — have argued that a structured survey inevitably lags the pace at which the technology is changing, and that some of the framings in the book have already been overtaken by the pace of Ukrainian and Russian adaptation. Analysts of a different stripe, closer to the arms-control community and to voices like Paul Scharre, have argued that the volume treats the ethical and legal questions with less fire than the material warrants, presenting them as one problem among many rather than as the central issue on which the whole edifice ought to be examined. Both critiques are recognisable as the standard traffic around any book of this kind, and neither dislodges the book’s basic contribution, which is to give practitioners and policy readers a single organised place to start.

For a reader working their way through the field, the volume sits in a specific slot. It is more comprehensive than Scharre’s Army of None, which is really a book about the ethics and near-future risks of autonomy rather than a survey of unmanned systems in current use, and it is more analytical than Hambling’s Swarm Troopers, which is closer to embedded reporting on quadcopters and their tactical implications. It is drier than Kenneth Payne’s I, Warbot and less concerned with the deep intellectual history of machine cognition; it is more current than earlier volumes like Peter Singer’s Wired for War, which now reads as a period piece. It pairs well with the more focused literature on the war in Ukraine, particularly the emerging body of work on the Ukrainian defence-innovation ecosystem and the Russian adaptation to loitering munitions, and it complements the more technical literature on machine learning in defence coming out of the RAND Corporation and CSIS. What it does not attempt, and does not pretend to attempt, is the ground-level account of what it is like to fly a drone into a Russian trench or to be on the receiving end of one; readers looking for that will need to go to journalism and memoir. It is unlikely to be the book anyone reads for the first time on this subject and very likely to be the book that ends up on the shelves of the people responsible for translating the subject into procurement decisions, doctrine documents, and export policy.

The parts of the book that seem likely to age well are the frames rather than the specific weapon names. The observation that a cheap, iteratively developed, software-defined weapon can hold an expensively-defended target at risk if it is produced in sufficient numbers is not going away; nor is the argument that a legacy procurement system oriented around small numbers of exquisite platforms cannot sustain the same fight. The industrial-base argument, the emphasis on middle-power exporters shaping regional balances, and the treatment of AI as a workaday tool inside the kill chain rather than a spectacle are the pieces of the book most likely to remain useful five years on. The parts that are already dated are the ones the pace of the current war has moved past — specific loss ratios, particular tactical adaptations, some of the counter-drone approaches that have been superseded by cheaper alternatives — and Rossiter and Layton are candid about the perishability of that material. That candour is one of the more honest features of the book. It closes with the recognition that a survey of a moving field cannot be a final word, and that its value lies less in the specific claims it makes about any one system than in the shape of the framework it lays down for readers who will keep watching after the covers are closed.