Two thousand machines, but above all two thousand customers to support
On 17 September 2026, ABZ Innovation inaugurated a €7 million expansion in Szentendre, Hungary, to manufacture heavy civilian drones. According to Reuters, the company expects capacity of 2,000 units a year on one shift and up to 3,500 on two shifts. Its chief executive, Karoly Ludvigh, said the expansion should increase production fivefold by the end of 2027. These are company targets, not volumes already delivered.
The number looks small beside consumer electronics, but these are different products. The machines carry liquids, seeds, fertilizer or equipment; they work near crops, buildings and people; and they require training, maintenance and permits. Reuters reports that ABZ sells in more than 40 countries and primarily targets the US and European markets. The company nevertheless ended 2025 with a loss of one billion forints, about $3.2 million, attributed to the costs of starting production, and hopes to break even within a few years. A factory therefore does not yet amount to a mature business.
What matters is the move from near artisanal assembly to repeatable capacity. An industrial drone must leave the line with traceable parts, consistent configurations and comparable test procedures. Only then can a distributor promise spare parts and repair times without relying each time on the engineers who assembled the first machine. Scale is useful for more than reducing costs: it turns a special machine into a working tool that can be bought and serviced.
More than a camera with propellers
ABZ's website presents a range from agricultural spraying and spreading drones to platforms for industrial cleaning, search and logistics. For its agricultural L50, the company claims a 50 litre tank, a maximum flow rate of 16 litres per minute and productivity of up to 24 hectares an hour. For the industrial C10, it specifies high pressure cleaning at heights of up to 60 metres. These figures come from the manufacturer: they describe claimed specifications and performance, not independent comparisons across every working condition.
The difference from a camera drone lies in the payload and the consequences of error. Thirty or fifty litres alter inertia, range and risk. A drop that is too large may fail to cover a plant; one too fine may drift away from the target. Wind, slopes, rows of vines, cables and trees make the route less predictable than an open field demonstration. The customer is buying neither flight nor spectacle: they are buying hectares treated at the right time, walls cleaned without scaffolding or supplies delivered where needed.
In February, Cadena SER reported an agreement with Spain's Fumidron for distribution in Spain, Portugal and Italy and trials in the Bierzo region. The distributor identified vineyards and orchards on uneven terrain as use cases and cited an average price near €20,000. These too are commercial statements reported by the newspaper, not a universal price list. The useful point is that, to reach the Italian market, the maker is doing more than selling online: it is building a local network that can demonstrate the machine, train pilots and provide after sales support.
The real competitor is the tractor the customer already owns
For a farm, the right comparison is not between a European and a Chinese drone selected from a catalogue. It is between the full drone service and the existing method: tractor, sprayer, operator, fuel, water, waiting times and potential crop damage. A drone becomes attractive when the ground is too wet for a vehicle, a slope makes driving dangerous or the treatment window is short. It does not automatically replace a tractor in every field, nor does it make the work free: batteries, pilots, maintenance, liquid preparation and transport still cost money.
ABZ says its agricultural range can reduce water use by 90 percent, chemicals by up to 50 percent and operating costs by 75 percent. These are promotional estimates from the company and need checking crop by crop, comparing dosages, agronomic effectiveness and weather conditions. Saving water does not necessarily mean a proportional reduction in active ingredients; treatment quality matters more than the number of litres carried.
This is where a market for specialist operators might arise. Many farms may prefer not to purchase and operate an aircraft costing tens of thousands of euros, but to pay for a service per hectare when needed. A provider can accumulate flight hours, spread the cost over several customers and keep trained staff. That is an economic hypothesis, not a guaranteed outcome: seasons, distances between fields and days when flying is impossible could leave a fleet idle while its instalments continue to fall due.
The invisible factory of permits and air traffic
Making more drones is insufficient if they cannot fly regularly. Reuters quotes consultant Gergely Kis, who says slow progress on unmanned traffic management systems is holding back European civilian applications. The European Commission calls U-space the set of digital services intended to coordinate numerous, automated operations and progressively integrate them with conventional air traffic. In its Drone Strategy 2.0, the Commission estimated that, given the right framework, the European drone services market could reach €14.5 billion and 145,000 jobs by 2030. That was a scenario projection published in 2022, not the market's current value.
U-space may include identification, flight authorization, traffic information and restrictions. It is software and regulatory infrastructure: invisible in a photograph of a drone, yet crucial to how many missions can be planned without phone calls, forms and manual coordination. For simple line of sight operations in an isolated field, a machine may work under relatively limited procedures. For flights beyond visual line of sight, deliveries or operations in complex areas, the ability to identify and separate traffic becomes part of the product.
The European strategy envisages widespread inspection, mapping, emergency and delivery services by 2030, with U-space fully deployed in several member states. The document also acknowledges the need for rules, standards and public acceptance. Privacy, noise, safety and liability are not external inconveniences: they are business requirements. A company that ignores these costs may demonstrate a successful flight without being able to repeat it at scale.
Where an Italian business could enter
The Hungarian news does not mean Italy should open another drone factory tomorrow. ABZ must first fill its announced capacity with orders and support those machines over time. More realistic opportunities may lie along the supply chain: distribution, training, spares, batteries, planning software, sensors, insurance and operational services for farming or maintenance. They are less spectacular than the aircraft, but could generate recurring revenue if they solve a measurable problem.
A concrete starting point would be to select one use case and measure it. In a vineyard: hectares treated, actual consumption, coverage, drift, preparation hours and total cost compared with the alternative. On a building: area cleaned, time, water, personnel exposed and downtime avoided. A demonstration should include failures and bad weather, not just the best flight. After a season, it becomes possible to see whether the customer is buying an advantage or a novelty.
European origin could also become a selling point, but it is not enough on its own. Parts availability, software updates, data handling and delivery times must be verifiable. Reuters reports that the expansion was funded by Assembly Ventures, Vsquared Ventures and Day One Capital, and that ABZ presents it as a way to reduce dependence on Chinese manufacturing. Yet the resilience of the supply chain depends on where motors, cells, semiconductors and sensors come from, details that the consulted sources do not provide.
The test is not takeoff, but the second year
The new Szentendre line makes a bet visible: Europe could sustain civilian production of heavy drones at a scale large enough to supply an international network. The first test will come from 2027 volumes, margins and the number of customers buying a second machine. The quieter test will involve spare parts availability, repair times, incidents, permits and cost per mission.
If these elements work together, the drone will stop being an extraordinary object and become a service machine, like a specialized tractor that uses the sky. If one link remains weak, the factory may produce faster than the market can put the machines to work. Here, the frontier is not making a machine fly. It is building everything that allows it to return to work the next day.
