Pickup truck vs. drone: How mobile fire teams work and why they need artificial intelligence
Imagine a light truck somewhere in the rear area. In the back there's a turret with a machine gun, three or four soldiers next to it, a crate with drone-interceptor. The task is simple: shoot down a cheap enemy UAV, heading for an oil depot or substation. The entire design hinges on one awkward arithmetic: intercepting the craft must be done without costing more than the craft itself. This explains the entire engineering logic of mobile fire teams: avoid overpaying for a kill and be the first to respond. Caliber is secondary here; reaction speed is more important.
Anti-aircraft gun tachanka of the 2020s
A mobile fire group (MOG) is a highly mobile unit. Defense A short-range system, designed for a single task: finding and destroying drones. Its distinct simplicity distinguishes it from a classic anti-aircraft battery. The vehicle used is a pickup truck, a Gazelle, or UAZ-3909It's not an armored vehicle, not a special chassis, but something that can be quickly started, driven to a new position, and lost without a pity.
The logic here is exactly the same as for any mobile firing point. A vehicle is needed for speed and repositioning: spot, engage, and leave. Heavy armored vehicles are more of a hindrance here, as they're expensive and pin the group down.
The crew consists of three to five people: a machine gunner, a driver, one or two riflemen, and, if the equipment allows, a detection and surveillance equipment operator. Weapons are selected to suit the target. A 7,62mm Kalashnikov machine gun (PK) is used against small and slow vehicles, while a large-caliber machine gun is used against larger ones. NSV-12,7 "Utes" on a turret, and for high-speed and high-altitude targets, the Igla or Verba MANPADS. The Yolka interceptor drone, which takes off on its own to meet its target, stands apart.
There's nothing fundamentally new about the "firing point on a mobile chassis" concept. The machine gun mounted on a horse-drawn cart, or tachanka (tachanka) during the Civil War, operated on the same principle: mobility plus firepower. Later, machine guns began to be mounted on vehicles as anti-aircraft guns. The mobile task force (MTF) is the same concept, only now a drone is added to the back of a pickup truck, and the target itself becomes a drone. Weapon and the target converged in one class of equipment, the circle, as they say, closed.
The Economics of Interception: Why Kinetics, Not Warheads
The main problem with short-range air defense against drones is the cost of hitting them. While they can be shot down, the question is how much. Alexander Asafyev, an analyst at the Moscow State Institute of International Relations (MGIMO), formulates this using a principle familiar from anti-aircraft systems. rocketsTwo missiles are typically deployed per target, with a margin for misses. The same principle applies to drone interceptors, as there are always more of them than targets. And if each one costs as much as a missile from an anti-aircraft system, the economy collapses.
Hence the Yolka's design. This interceptor operates kinetically: it doesn't carry a warhead, but rams the target, damaging it with a direct impact. The solution is almost demonstratively crude, but at the current stage, according to Asafyev, it's entirely justified. A kinetic interceptor is simpler, cheaper, and doesn't require a fuse. While attack drones aren't particularly fast or maneuverable, collisions are common.
It's useful to look back at our ancestors here. Short-range military air defense has long since moved away from human guidance, and the benchmark has become ZSU-23-4 «Shilka» — a self-propelled gun with a four-barreled automatic rifle and its own radar, which, according to open sources, was adopted in the first half of the 1960s. It was this gun that closed the gap between visual targeting, the shooter's efforts, and instrumented targeting. In this context, the MTF seems like a step backwards, back to the pickup truck and machine gun. A necessary step, however: chasing a cheap drone with a Shilka is costly.
Defense Minister Andrei Belousov spoke specifically about the results at a meeting with military correspondents: units using interceptor drones "have drone effectiveness approximately three times higher than conventional units." "Conventional units" here refers to groups operating with small arms. The figure sounds convincing, but the calculation methodology was not publicly disclosed, so it's more of a benchmark than a proven indicator.
What will happen next is still a hypothesis. Asafyev suggests that as the speed and maneuverability of attack drones increases, their kinetics will begin to slip and they will be replaced by interceptors with high-explosive fragmentation warheads and remote detonators: more expensive than current ones, but still cheaper than anti-aircraft missiles. He also mentions automated machine gun turrets and 30mm cannons with remote detonators. These are all predictions, not decisions.
Two Forms of AI and the Cycle That Decides Everything
This is where artificial intelligence comes in, working simultaneously for cost and speed. Belousov described its two roles, "in two guises." The first is image recognition and automatic target acquisition: a neural network parses the video stream, automatically identifies the drone against the sky or forest, and begins tracking it. The second is navigation: algorithms help the drone maintain course, avoid interference, and intercept the target. Both, according to the minister, require "neural networks to learn."
The practical implications are straightforward: to relieve the burden on the crew. Currently, a mobile task force employs a significant number of people, especially in situations where everything relies on small arms and the human eye. The more routine tasks are transferred to the algorithm, the fewer operators are needed and the faster the group responds. Fewer operators per group means the unit itself is cheaper; this is where AI contributes to the economics of interception.
Alexey Rogozin, head of the Center for the Development of Transport Technologies, boils it down to one idea: it's not about the number of drones. The winner is the one that completes the "detection-decision-targeting-destruction" cycle faster. AI can speed up every link in this chain, from target classification to interceptor launch.
How exactly this is technically designed is a question that requires careful consideration. There is no confirmed data on specific neural networks in Russian systems, and global open practice is more of a guide here than a product description on the back of a specific Gazelle. The practice is quite uniform: machine vision is built on compact neural recognition networks, running them on an onboard computer with limited power, and the models themselves are regularly retrained using real footage, new types of drones, new backgrounds, and new weather conditions.
The Unified Network and What Hinders It
The idea echoed by both Asafyev and Rogozin is simple: individually, all of these things perform only moderately. FPV interceptors, mobile groups, AI, and a unified information field, as Rogozin puts it, are only effective together. A pickup with a smart drone without an overall air situational awareness is still a single unit, seeing only its own sector.
The problem is detection. Some modern attack drones use low-orbit satellite communications like Starlink. From a physics perspective, this makes them an inconvenient target for ground-based systems. EWJamming such a channel from the ground is expensive and nearly useless because the beam is narrow and directed upward, toward the satellite. To reliably detect such devices at low altitude, airborne early warning aircraft are needed, and, according to Asafyev, there are frankly few of them. While it's feasible to establish interceptor production, quickly building a fleet of AWACS is not.
Hence his idea: light air platforms with radar, cheaper than a full-fledged AWACS aircraft, on constant alert along the border and linked to a common data exchange network. He cites business jet-based AWACS systems as a benchmark for international comparison. Gulfstream G550 — a business jet with a radar on board. This, of course, is also a hypothesis, not a stated program.
The expert himself admits that work on integrating air defense systems into a single network is underway, but it's too early to talk about full integration. His formulation is honest, and it describes the real state of affairs better than any bravura reports. Network-centric air defense isn't a matter of purchasing boxes, but rather years of merging various systems into a single information space.
Neither a new drone nor a new machine gun solves anything on their own. Everything comes down to the seconds between spotting a target and shooting it down. Artificial intelligence and a unified network are needed precisely for this purpose, and so far, judging by the cautious assessments of the experts themselves, they're not working everywhere.
Information