Automation Does Not Lead to Leaner Land Forces

3/4 conducts urban operations with unmanned systems at Range 220 during Exercise Apollo Shield

In June 2022, I travelled with a Ukrainian reconnaissance unit that was seeking to get an uncrewed aerial system with a camera over a series of Russian positions. Although the drone they were using was autonomous, this was not achieved from a remote headquarters with the press of buttons.

Instead, a convoy of vehicles and eight personnel had to thread their way into no-man’s land, establish a communications array, and hold their position for several hours. I was reminded of this by the juxtaposition between the portrayal of drone operations in Pentagon PowerPoint slides and their reality as observed in Ukraine.

It has become fashionable — in the face of recruitment challenges across Western militaries — for military leaders to assert that the impact of falling troop numbers is mitigated by the declining requirement for mass, owing to the promise offered by robotic and autonomous systems being introduced into the force. The problem with this argument is that, as far as land forces are concerned, it is entirely without evidence to justify it. As I have seen in Ukraine and have observed in other theaters, the introduction of robotic and autonomous systems into the force is liable to increase both the number of people and the diversity of skills necessary within the force.



The drone needed that many people to support it because the mission necessitated the operator, a technician, and a communications specialist, as well as the force protection to keep them alive while they were doing their job. The technology was sophisticated, but that did not stop it from being labor- and skills-intensive. Nor were the personnel requirements limited to the people in the field. To plan the drone’s flight path, electronic warfare operators were required to provide an electromagnetic survey, and to exploit the images captured, it was necessary to have image analysts. The reality of most emerging technology is that it requires people, and if the number of soldiers is reduced in one area, they are often displaced to other parts of the battlefield.

Irreducible Minimums

Although autonomous systems can often magnify the capabilities of a unit, there are irreducible minimums in warfare as to how many people are needed to perform basic tasks. Consider an infantry squad that is furnished with a small autonomous tracked vehicle carrying a machine gun and a sensor payload, as well as a small drone for surveillance. Let’s assume that the size of an infantry squad without these tools is normally 12 soldiers, divided into three fire teams. For the squad with the robotic and autonomous system, It is clear that two of the fire teams must remain unchanged. This is because autonomous systems have clear limitations. A tracked vehicle cannot assault a trench, or climb into a building, or process prisoners. Thus, the infantry squad still requires two assault fire teams.

The third fire team, comprising the base of fire element, could make considerable use of the robotic and autonomous systems. In testing, these systems have demonstrated that they can deliver far more effective, responsive, and accurate suppression than a human machine gunner. However, they cannot be left to do this on their own. Robotic and autonomous systems are poor at judging context, and without a human on the loop (but not in it), I have watched them do all manner of stupid things, from continuing to shoot killed targets to becoming stuck by such simple obstacles as a farm gate.

The problem is that once one person is required to remain with the robotic and autonomous systems, it becomes necessary for a second to be with them. It is not a good idea to have someone left by themselves on a battlefield. If they are injured, for example, someone needs to be nearby to support them. Being left alone while facing an immediate threat to life is also liable to impact morale and decision-making.

Then there is the issue with uncrewed aerial systems. These drones are becoming critical to tactical actions, as I have personally seen in my many visits to Ukraine. But the operator using a drone is looking into a pair of goggles or at a control terminal. The operator lacks any wider situational awareness. It is inappropriate for them to be with the assault teams, and it is unwise for them to be left by themselves while operating in close proximity to the enemy. Thus, the squad with the robotic and autonomous systems and drone might be much more lethal than the squad without them, but it does not necessarily get any smaller.

The Tail Counts

There are some manpower efficiency gains if novel technologies are concentrated rather than distributed out to squads. For example, it is entirely feasible for two personnel to oversee the engagements of an entire antitank platoon’s worth of robotic and autonomous systems equipped with antitank guided missiles, or to manage the engagements of a robot machine-gun platoon. In the context of a battalion fire support company, therefore, one might see fewer crew per crew-served weapon. The effect, however, is not to reduce the number of personnel required, but to displace them into the support company.

While two personnel might oversee an emplaced antitank missile screen, there are a lot of tasks, including keeping watches while sleeping, that demand more people. But the real kicker comes from the maintenance burden of the robotic and autonomous systems. A traditional machine-gun platoon in an infantry battalion is carrying relatively simple mechanical objects that they can keep running, with the assistance of a battalion armorer and their tools.

A robotic and autonomous system–enabled machine-gun platoon would have to manage a fleet of complex vehicles comprising mechanical drive trains and power packs, sophisticated electronics in their sensors, the weapons they carry, and the software and communications equipment that makes them functional. Moreover, when a vehicle like this suffers damage to its drive train, for example, it must be recovered. It is too heavy to be lifted. And so the battalion now needs a vehicle capable of towing the robotic system.

I can distinctly remember instances during trials on Salisbury Plain and in the United States where autonomous vehicles navigated themselves into ground that was too soft to support them and became immobilized . The recovery vehicle also needs operators and must be maintained. Nor is it reasonable to presume that the same fitter who manages the track and power pack will necessarily have the expertise to be able to reprogram, update, debug, or manage the software governing the system. There is a need not just for comparable numbers of people, therefore, but also for an expanded skill set, including some that are in high demand in the civilian economy.

Another factor that is rarely considered when the personnel efficiencies of robotic and autonomous systems are propounded is that more capable platforms often come with opportunities that require personnel to exploit. A good example of this are Predator and Reaper drones. Initially, they appeared to save personnel. While they required two crew — a pilot and payload operator — and often three or four crews, given the duration of their missions, compared to the number of aircraft that would need to be flown to provide comparable sustained coverage, and without the need for personnel on standby to recover downed pilots, these appeared to offer efficiency gains.

But then it was noticed that while the crew might be watching a specific point through an MX-15 sensor ball, it was possible for the platform to harvest imagery across a much wider area during missions and to collect other things, like signals intelligence. Actually, making sense of all of the data collected, however, came to depend upon a large number of intelligence analysts and technicians. Thus, by the 2010s, the number of people involved in a given Reaper orbit could be as many as 150. Further automation of analysis has since reduced this number, but the size of the exploitation force remains significantly greater than what had previously supported aerial reconnaissance.

The final problem with the idea that automation and technology will drive a reduction in personnel is that these visions are rarely presented in an adversarial context, where the enemy is trying to disrupt these capabilities. For example, the ability of drones to deliver mass precision strike to tactical echelons could be a significant capability. However, it is only likely to be effective if supported by the ability to conduct electromagnetic surveys and other intelligence preparation to plan strikes, so as not to be countered by enemy electronic warfare. Conversely, on a battlefield where the enemy fields such capabilities, the need for short-ranged air defense and organic electronic protection capabilities across tactical formations introduces the need for additional specialists in a combined arms unit that were not previously needed. The result is a redistribution rather than a reduction in headcount.


The military — often with justification — sees advantage in new technology. There is also the hope that new technologies will solve today’s frictions. In other words, there are multiple factors that encourage a cognitive bias in Army leaders to believe that the challenge of recruitment can be innovated around.

But when it comes to people, technology is labor-intensive. There are irreducible minimums of people to provide the assured capacity to conduct battlefield tasks. Often, technology displaces people but does not eliminate them. It is feasible today for a handful of people to orchestrate a survey of a vast array of sensors and assign a broad selection of weapons to targets — such as with an integrated air defense network — but it still takes many more people to keep that network integrated. Perhaps most important, exploiting the advantage that technology offers often creates requirements for additional personnel performing tasks that were not previously envisaged.

In this context, it is vital that militaries address the recruitment challenge by assessing the offer they are making and adapting it to ensure they attract the people they need. This is especially important if those with skills that are also in demand in the civilian economy are to be brought into and kept in uniform.


Dr. Jack Watling is senior research fellow for land warfare at the Royal United Services Institute.

Image: U.S. Marine Corps photo by Lance Cpl. Justin J. Marty