Franz-Stefan Gady, Alexander Stronell
Last week’s Russia-brokered agreement between Armenia and Azerbaijan ended 44 days of bloody clashes over the disputed enclave of Nagorno-Karabakh—the first interstate war fought by conventional forces in recent years. The deal calls for Armenia to give up large swathes of territory in and around the breakaway region, which lies within Azerbaijan’s internationally recognized borders, and Armenian Prime Minister Nikol Pashinyan called the deal “incredibly painful.” The ostensible Azerbaijani victory, gained at substantial cost in men and materiel, has triggered intensive interest among military analysts about the conflict’s lessons for future warfighting.
In particular, the wearing down of Armenian air defenses and armored platforms by unmanned aerial vehicles, or UAVs, has led to significant debate over the continued utility of main battle tanks for high-intensity military operations. Indeed, a narrative has emerged in the media, fueled by propagandized videos shared on social media channels, that armor may be obsolete in the face of air-launched precision strikes carried out by UAVs.
This conclusion, however, obscures the durable and important tactical lessons to be learned from the conflict, of which seven major takeaways stand out.
First, as other analysts have pointed out, the fighting does not offer any conclusive evidence that battle tanks and other armored vehicles no longer serve a purpose in the modern battlespace because of UAVs. To paraphrase Winston Churchill, armor continues to be the worst form of protection for soldiers against UAVs—except for all the others. Military personnel who were caught during the recent fighting in the open or in fixed positions, as well as in command posts and unarmored transports, fared considerably worse than soldiers who were struck while in armored vehicles, in part due to the limited armor-penetrating payload that UAVs can carry. By the same token, there is still no substitute for tanks and other armored vehicles for maneuver warfare and as close-range fire support for advancing infantry in the modern battlespace. UAVs and lightly armed infantry are ultimately not capable of holding or defending ground for a prolonged period of time.
Second, the methodical use of air-launched precision strikes by Azerbaijani forces to systematically degrade Armenian air defense assets confirms the importance of a large stock of short- to long-range kinetic strike munitions paired with dispersed intelligence, surveillance, target acquisition and reconnaissance, or ISTAR, capabilities. UAVs can serve as useful and expendable strike platforms in that regard, and European militaries should accelerate their planned acquisitions of them. However, this will likely be a tall order for small and medium-sized military powers that will only have the financial resources to acquire the necessary munitions and ISTAR assets in small numbers. Consequently, it can be reasonably expected that in future conventional wars, advanced ISTAR and precision-strike capabilities will only be used at the outset of a conflict, since the former will be quickly disabled and the latter quickly depleted.
Third, the conflict has also shown the importance of non-kinetic long-range strike assets, particularly electronic warfare, or EW capabilities, to defend against both kinetic and non-kinetic strikes. Azerbaijan quickly took Armenia’s isolated EW assets out of action in the first days of the war. Thereafter, Azerbaijani UAVs operated outside the range of Armenian tactical air defenses or often flew too low to be engaged by Russian-made systems like the short-to-medium range Pantsir-S1. Armenia could have likely countered at least some Azerbaijani UAV operations with the deployment of integrated EW assets such as long-range jammers, as the unconfirmed but reportedly successful deployment of a Russian EW system demonstrated. EW alone, however, is no panacea and needs to be integrated with other air defense systems to effectively defend against aerial threats.
Fourth, the conflict has shown the need for novel operational concepts and doctrine that merge legacy platforms like battle tanks and electronic warfare assets with emerging technological capabilities, including UAVs. This also includes the integration of defensive and offensive cyber operations and information warfare. Although cyber operations do not appear to have played a major role in this conflict, they will likely be utilized to disable or even destroy command-and-control assets and air defense systems at the outset of a future high-intensity conflict.
The key military resource in the next decades of warfighting will remain the human operator.
In the long run, all advanced militaries will have to make the transition from a platform-centric to a more integrated network-centric force structure. In the decades ahead, the military that best understands how to combine the two into a hybrid force structure that integrates legacy systems with emerging technologies will possess the advantage in high-intensity combat. This applies to small, medium and great powers. A lot of legacy materiel is going to stay around, and militaries need to find ways to upgrade and integrate it.
Away from the frontline, the conflict has undoubtedly confirmed the importance of the information space to future conflict. Both sides expended significant resources attempting to control the narrative surrounding the fighting, mobilize international public opinion in their country’s favor and attract the attention of foreign governments. Social media platforms, particularly Twitter, were the central arena for information warfare efforts, with ordinary citizens and bots alike sparring in so-called hashtag narratives, disputing the outcomes of fighting in Nagorno-Karabakh and alleging—at times with justification—aggression and atrocities on both sides.
Fifth, the conflict has shown both the ever-increasing importance in modern conflicts of effective, synchronized, combined arms operations at the tactical level and the exponentially more devastating consequences of failing to carry out such operations. Once a single component in a combined arms operation fails—such as infantry failing to support armor, or air defense assets not protecting artillery—it becomes easier to punch a hole in the defense and quickly exploit it. For example, Armenia sustained exceptionally high battle tank casualties during a counteroffensive in the first days of the conflict, after Azerbaijan succeeded in destroying over a dozen short-range surface-to-air missile systems. New anti-aircraft systems can only offer a partial remedy unless and until they are integrated into a wider sensor network and linked with other air defenses. This is even more true once militaries begin to effectively deploy swarms of UAVs or remotely controlled ground vehicles. At the same time, given additional vulnerabilities in space and cyberspace, combined arms operations have been made more complex.
Sixth, the high attrition rate of materiel in the conflict points to the need for platforms and weapons systems that are more “attritable,” or disposable and requiring less maintenance. But it also establishes that, without proper training in combined arms operations undergirded by a solid mission-command doctrine that emphasizes tactical flexibility, forces are bound to suffer an exceptionally high attrition rate in modern conventional warfare. For example, a preliminary analysis suggests that Armenian forces were too slow to adapt and alter standard operating procedures and doctrine when it comes to supply and logistics operations. The same goes for troop assembly areas, which at the beginning of the conflict were too close to the frontline and largely unprotected. (Armenian fixed positions were also by and large not hardened to offer protection against aerial and artillery strikes, demonstrating the need for renewed doctrinal emphasis on military engineering, in particular combat engineering.) Furthermore, as the military analyst Rob Lee pointed out, the conflict has shown the need to empower junior commanders at the company and platoon level to conduct maneuver in smaller formations.
This leads to the seventh and perhaps most important observation about the conflict: The key military resource in the next decades of warfighting will remain the human operator. As Robert Bateman highlighted in explaining the high attrition rate for tanks in the conflict, the ubiquitous videos from the first days of the fighting ostensibly illustrating the lethality of UAVs actually showed armored vehicles “clumped together in tight clusters … not maneuvering while dispersed widely as the conditions in combat would warrant.” This points to human error caused by insufficient training. Lower casualties on the Azerbaijani side were not just the result of superior UAV capabilities, but also point to better preparation and training. This should not come as a surprise given that Azerbaijan chose to initiate hostilities.
Consequently, the war underlines the need for additional investments in real-life military training that emphasizes dispersed combined arms operations or its future iteration: multidomain operations. In the future, military commanders may have all the strike capabilities and AI-enabled predictive analysis in the world, but if they lack fluency in the fundamental principles of maneuver warfare, they will likely still suffer unnecessary casualties in the future battlespace.
Finally, a word of caution is in order. As with all conflicts, the Nagorno-Karabakh war had unique attributes that make generalizations difficult, and there is a risk that we draw the wrong conclusions from the fighting when it comes to great power, high-intensity warfare. However, if there is one enduring lesson cutting across all military conflicts, it is that a well-trained and well-maintained military beats ill-disciplined troops anytime, no matter the technological edge the latter enjoys.
Franz-Stefan Gady is a research fellow at the International Institute for Strategic Studies (IISS) focused on future conflict and the future of war. He tweets @hoanssolo.
Alexander Stronell is a Russian-speaking research assistant based in IISS’s London offices working on cyber projects. He tweets @AEStronell.