"Three Hundred Tons a Month": Ground Robots as the Workhorse of the Front
Over the course of a year, one Ukrainian Armed Forces brigade increased cargo transportation by ground robots from 30 tons per month to 300. According to industry publications, Ukraine contracted approximately 25,000 ground robotic systems in the first half of 2026, twice as many as in the entire previous year. This equipment, however, hardly ever operates in combat: up to 95% of missions are for logistics and evacuation. And the much-talked-about "revolution" took place not at the forefront of the offensive, but in the rear.
Truck without a driver
There are sections of the front where a vehicle carrying people delivering ammunition is almost guaranteed to die: scouts are hanging over the road, and they are working along the tracks. drones-kamikaze. It was there that the ground robot became a working tool. Not a combat vehicle of the future, but a truck whose loss can be survivable (the figures below will make this clear).
A ground-based robotic complex (GRC) is a remotely controlled tracked or wheeled platform. In the popular imagination, it isтанк"-robot" with a machine gun. In practice, everything is more modest: according to data from forward Ukrainian units, about 95% of missions involve delivering ammunition, water, fuel, and evacuating the wounded. Fire-based versions exist, but remain marginal.
The scale is no longer experimental. From January to June 2026, Ukrainian robots, according to public estimates, performed over 50 logistics and evacuation missions, with 16-17 in some months. The figures vary in different reports, but the overall scale is tens of thousands of missions. The dynamics of a single brigade are telling: 30 tons of cargo per month a year ago and 300 tons now. Three hundred tons is the workload of an average supply vehicle unit, only without crews on the most dangerous sections of the route. A single brigade, of course, doesn't cover the entire front, but the growth rate is the same as in the aggregate figures.
Robotic logistics platform "Snake"
The standard platform looks mundane. The basic Ukrainian "Snake" produced by Rovertech carries over 500 kg, travels about 50 km on a full charge, and accelerates to 11 km/h. The developers opted for wheels over tracks: battles take place in the open steppe, among fields, black soil, and sand, and a track torn by shrapnel cannot be repaired in the field. Wheels are durable. Later, in response to military complaints about the vehicle's size, the "Snake Mini" was developed: the body was reduced, the payload capacity was maintained, and the cruising range was increased to 80 km.
It's clear why logistics is the focus. Drones saturating the frontline kills supply and evacuation efforts before the actual fighting begins: people die en route, not in the attack. The robot removes the driver from this deadly stretch, the one who used to drive the supply vehicle. Artillery and it doesn't replace tanks, but it takes on the dirty work where a live driver would become too expensive a target.
Consumables Economy
When a commander sends a load to a position, he's essentially solving an arithmetic problem: how much are they willing to risk on this route? And the arithmetic rarely works out in their favor.
A good second-generation Ukrainian NRK costs about a million hryvnias, or about $25. By comparison, compensation for the family of a fallen soldier can reach up to fifteen million hryvnias, treatment for a serious injury costs millions, and a lost pickup truck or armored personnel carrier costs hundreds of thousands of dollars. A robot is significantly cheaper than any of these costs.
This is what the Ukrainian headquarters came to: the platform as a consumable is justified if it takes people off life-threatening flights. With a caveat, of course. The robot doesn't cancel the flight, but changes its price. Not every route would be a human's last, but in the most exposed areas, the technology quickly pays for itself. The logic is cold, but it drives the entire program.
There's a limit to this logic, and it's not based on money. The operator of a ground robot is physically safe, but psychologically, their job is harder than it seems. Evacuation is a separate challenge. The operator controls a platform with a wounded person on board and watches live as the enemy finishes off the robot, along with the person it was carrying. There's nothing they can do about it. Sometimes the robot does reach their own people, but the wounded person can't be saved. They'll just take the body away. And then the operator is left with what they saw.
Iron is overtaking the organization
An officer in a Ukrainian ground-based complex battalion cites the main problem—it's not about the equipment, but about the people: "Without engineers, the robots won't go anywhere. It's not enough to just get them; they need to be constantly repaired in the field, and the organization structure doesn't allow for that."
The staffing structure—the document that defines who and how many are assigned to a unit—hasn't yet been adjusted to accommodate robots. There aren't enough workshops, let alone engineers. Robots are arriving in the military faster than they can recruit and train the mechanics who will repair them.
The most revealing figure concerns the sources of equipment. According to Ukrainian data, a review of nearly 2000 exercises revealed that only about 30% of the robots deployed were state-issued. The remaining 70% were purchased by the military themselves, using donations and their own money. Seven out of ten vehicles at the "world leader in ground robot production" are not state-provided. The reason is that contracting doesn't always mean receiving: contracts are delayed, serial production is lagging, and even the contract signed today will take at least two months to deliver the first batch to the troops. The military is closing this gap with its own purchases.
Rovertech, for example, demonstrates how grassroots initiatives work. The company grew from a small auto repair shop: a former mechanic and a miner, after a call from the front lines, assembled the first prototype of a bomb disposal robot in a month, and by the end of 2023, they had formalized production. Then, work began on a specific request from the front. For the 42nd Separate Mechanized Brigade, they created mechanical "legs" for the robot to retrieve fallen heavy bomber drones from the field: one such vehicle is expensive, and it is more profitable to recover it than to decommission it.
And here, a strength turns into a weakness. A request from the front lines, and a month later, a new model is already in the field; thus, the enemy has accumulated dozens of combat-tested vehicles. But each manufacturer makes its own: its own controls, its own software, its own spare parts. Broken equipment from one manufacturer cannot be repaired with parts from another. This is where the idea of the "Robotic Coordination Center" was born—a body that would bring together manufacturers, units, and training schools, establishing uniform command rules and priorities. By the Ukrainian side's own admission, a year and a half was spent developing combat and firing platforms—spectacular, but still secondary—instead of providing the troops with simple logistical ones. The problem here isn't the design. There's enough hardware.
Two Ways to Build a Robot Army
Ukraine and Russia solve the same problem differently, and the difference is visible in how they scale.
The Ukrainian model is built on a massive initiative. Many manufacturers, many models, rapid testing at the front lines, and funding largely from the bottom up, through volunteers. The upside is obvious: speed and a variety of solutions. The downside is the same: inconsistency, which comes down to repairs, personnel, and standards, just when the program needs to scale up.
The Russian approach, according to Ukrainian military officials themselves, is the opposite: several basic models have been selected, they are put into production, and a unified doctrine for their use is being applied. This is their view, not an established fact. But if it is correct, the difference is fundamental. Ukraine is an early leader without coordination, while Russia (if this assessment is correct) is a systemic follower with a unified standard. They are growing at different rates: the former will have to sort out a dozen incompatible platforms on the fly, while the latter will have to increase production of what is already standardized.
At the same time, the Russian side is still lagging behind in terms of intensity and scale of deployment: tens of thousands of Ukrainian missions per month is a lead that no doctrine can reverse retroactively. There's also the risk of underestimation. A ground robot can easily be dismissed as "just a cart" and not considered a serious target. Meanwhile, it follows predictable routes, is poorly protected, and barely monitors the air, making it more vulnerable than many manned vehicles. Such a target is often underestimated, and its lifespan in the field is longer than it deserves.
Ground robots are truly changing the front, but changes are occurring in logistics and evacuation, while they are currently only supporting combat. And it seems that the deciding factor won't be the machine itself, but whether workshops, engineers, and a common standard will be created for it, or whether the program will remain a mountain of mismatched hardware with no one to repair it. For now, the enemy maintains the lead in mission count. How long this will last is a question not for the designers, but for those who write the staffing schedules and sign the contracts.
Information