The sales pitch for ground robots is simple and seductive. Load a trailer with wheeled and tracked UGVs, send them into the dangerous bits of the battlefield, and come home with fewer casualties and fewer boots on the ground. That pitch is not wrong, but it is incomplete. In practice the early wave of deployed ground robots has a hidden manpower tax. You do not just replace a human with a chassis and a battery. You add operators, observers, maintainers, cyber and logistics teams, and new layers of command and control to keep those machines useful.
Field reports from Ukraine and open reporting from Western analysts make the basic problem clear. Most current UGVs are not independent agents; they are remote tools that need guidance, overwatch, and protection from airborne sensors and electronic attack. That tether to human control means a vehicle that looks cheap on a balance sheet often generates multiple personnel demands on the ground: teleoperators to drive and fight, drone crews to provide overwatch, teams to recover and repair, and staff to manage batteries, communications and countermeasures. The operational picture is not a simple one-for-one substitution. Deploying a fleet of ground robots can be manpower-intensive.
Human factors research explains why that is not just a wartime improvisation but a structural constraint. Studies of unmanned systems across domains show clear limits to how many vehicles a single operator can control or even supervise before situational awareness and performance degrade. One frequently cited finding from multi-phase operator studies is that with moderate automation one person can supervise many air vehicles but can typically control only a few when mission-level decisions and payload management are required. Similar multi-robot interface work shows that operator workload, attention bottlenecks, and degraded situational awareness become real problems as the number of remotely operated assets climbs. In short, more machines do not automatically mean fewer humans - they often mean different humans and concentrated cognitive load.
On the ground the problem compounds. Ground robots operate in messy, GPS-denied, obstacle-rich environments that strain perception algorithms and demand frequent human intervention. Teleoperation in urban rubble or forested terrain requires high-resolution video, low-latency control links, and occasionally a secondary visual assistant or remote scout to give the operator perspectives the UGV cannot. Those helpers are often additional robots or human observers. There are no universal plug-and-play solutions yet that let a single operator shepherd dozens of UGVs across complex terrain without unacceptable risk to the mission. The practical upshot is that the manpower cost per effective UGV remains high until autonomy and interfaces improve.
Combat experience reinforces the laboratory findings. Reporting from the Russia-Ukraine conflict shows extensive use of ground robots for logistics, demining and even weaponized tasks but also documents the platforms vulnerability to aerial observation, electronic warfare and the need for on-site support. Commanders have described UGV employment as valuable but manpower and protection intensive. When you add the routine tasks of charging batteries, rearming, field cannibalization for spare parts, and recovery under fire, a deployed UGV fleet becomes a maintenance and personnel burden as well as a tactical capability. That is not a failure of concept; it is a maturity problem for a technology moving from prototypes to mass operational use.
The military procurement impulse is often to buy more platforms to get the advantage of volume. That reflex can make sense if the intent is attrition and massed cheap systems. But if the objective is to preserve scarce trained soldiers by offloading dangerous tasks to robots, planners must count the entire support tail. The sustainment burden includes trained teleoperators, interface and human-machine team training, spare parts pipelines, secure and resilient communications, electronic protection suites, and doctrine that integrates UGVs with aviation and fires. All of those elements cost people and money. Ignore them and you end up with boxes that sit idle or become liabilities on the battlefield.
That said, the problem has tractable engineering and organizational mitigations. First, improve the human-robot interface rather than just the chassis. Immersive tools, predictive displays, and decision aids demonstrably reduce workload and raise situational awareness. Where possible, use marsupial teams - pairing an autonomous visual assistant or small UAS with a UGV - to reduce the need for a second human operator. Those tactics shrink the human-to-robot ratio without pretending autonomy has already solved perception and decision-making.
Second, design for logistics from day one. That means modular batteries and payloads, standardized spare parts, and maintenance-forward contracts that accept the reality of field cannibalization. It also means measuring systems not just on unit cost and headline range but on mean time between repair in contested conditions and the personnel hours required per operational hour. Procurement metrics that ignore sustainment create incentives to buy flashy prototypes that increase work for soldiers rather than reduce it.
Third, invest in autonomy where it matters and accept incremental delegation. Full autonomy in complex, lethal environments is still a research frontier. Focus instead on reliable sub-skills - robust navigation in rubble, obstacle negotiation, automated battery management, and constrained mission execution with conservative safety envelopes. Those capabilities reduce the frequency of human intervention and the attention budget required from operators. But autonomy must be introduced with clear human oversight and recovery procedures.
Finally, update doctrine and training. UGV employment demands new crew concepts that combine vehicle drivers, remote weapons operators, robotic maintainers, and cyber defenders. Realistic exercises that stress the whole pipeline - from resupply to repair to contested comms - reveal the true personnel cost much better than static demonstrations. If the goal is fewer human casualties rather than fewer humans on payroll, then integrated training and doctrine must be part of every acquisition plan.
The headline should be simple. Ground robots can reduce risk to soldiers, but they do not automatically shrink the human footprint. At current levels of autonomy and battlefield complexity, they often reallocate labor into concentrated teams that control, protect and sustain those machines. Recognize that trade-off up front. Buy the interfaces and logistics as aggressively as you buy the platforms. Train the people who will run the robots in the contexts they will actually face. That is how you turn the promise of ground robots into operational advantage instead of a surprise manpower bill.