NGRNWhat 2026 Is Likely to Reward
Overview by NGRN
By late 2025, Ukraine’s defence-technology ecosystem looks less like an improvisational wartime workshop and more like a fast-growing industrial sector under extreme operational pressure. The year’s defining feature was not a single “wonder weapon”, but a steady shift in what matters: resilience over elegance, integration over standalone gadgets, and software-defined adaptation over hardware-centric iteration.
2025 in Review: What Actually Moved the Needle
1) Drones: fibre-optic FPV, interceptors, and the end of “radio as default”
If 2024 was about scale, 2025 was about surviving the countermeasures that scale provoked. The single most consequential drone development highlighted by practitioners was the rapid mass adoption of fibre-optic FPV. First seen in Russian use in late 2024, the technology reached “mass-level” fielding in Ukraine by summer 2025, with some units reportedly moving away from radio links altogether.
Alongside that came the normalisation of “drones versus drones”: quadcopter-based interceptors used against one-way attack UAVs and other aerial targets, increasingly packaged not as lone aircraft but as mobile комплекси (systems) built for repeated deployment.
Yet the same interviews strike a sober note on what didn’t scale. Terminal homing/“last-mile” guidance remained unevenly fielded; and there was still no widely evidenced, combat-confirmed employment of true drone swarms (as opposed to coordinated multi-UAV tactics).
2) Algorithms over airframes: autonomy as a workaround for EW
Several teams described 2025 as the year autonomy moved from a conference slide to production lines — not full autonomy, but practical automation that can carry a drone through the final segment when control links are degraded. One engineering team described 2025 as a transition “from trials to scaling”, including mass production of drones with autonomous terminal guidance and deployment of dedicated anti–one-way-attack detection modules already used by Ukraine’s forces.
The underlying logic is blunt: EW made the last mile decisive, so the side that can keep the kill chain functioning under jamming wins disproportionate advantage.
3) The structural bottlenecks: standards, certification, and talent
Across the interviews, a recurring complaint is that parts of the requirements and certification culture were inherited from civilian aviation and do not match battlefield iteration cycles. The sector is simply moving faster than the rulebook — producing a growing gap between “demo reality” and combat reality.
Meanwhile, the labour market is tightening around a few scarce profiles — especially embedded developers and system engineers with guidance/navigation/networking depth.
That constraint is not uniquely Ukrainian: the EU itself is discussing industrial-scale reskilling, with a target to upskill/reskill 600,000 people for the defence industry by 2030, explicitly framed as a remedy for a skilled-labour bottleneck in rearmament.
Electronic Warfare: Why 2026 Will Be a “Physics Year”
EW is treated in the DOU material as an arena of slower cycles and harder science. One CEO argues there were no dramatic EW “breakthroughs” in 2025 because genuine capability shifts take years — and because component ecosystems (RF chips, amplifiers, ADC/DAC integration) govern what can be built and fielded.
The most strategically charged forecast is the rise of GaN-based high-power microwave amplifiers enabling more energy-efficient EW, potentially including effects against LEO satellite communications — with the text explicitly noting Starlink as the reference problem-set and suggesting first commercial solutions could appear by mid-to-late 2026 if the technology matures as expected.
The implication is uncomfortable but clear: in 2026, “clever” will matter less than electrodynamics, thermal design, and RF engineering discipline, because these determine whether a system works in harsh environments and at relevant frequencies.
Ground Robotics: From Curiosity to Logistics Backbone
On UGVs, the story of 2025 is credibility. What was previously treated with scepticism is now described as used in units for frontline logistics and evacuation — less glamorous than strike missions, but operationally transformative.
The 2026 direction of travel, as framed by robotics specialists, is evolutionary rather than revolutionary: limited autonomy, partial automation for difficult terrain and unstable comms, and a migration from “home-built” vehicles towards scalable platforms based on serial hardware (pick-ups, ATVs, heavier carriers).
Most telling is the push toward multi-domain teaming — UGVs operating as carriers and “mothership” platforms for FPV, paired with aerial drones for faster target work. This is less about robots replacing soldiers and more about robots reshaping how units move, resupply, and project effects under persistent observation.
Components, Comms and Software: The Quiet War That Decides the Loud One
If drones are the visible edge of the fight, components and connectivity are the decisive substrate.
1) GNSS-denied as the baseline
Multiple contributors describe a reality where GPS-denied conditions are no longer exceptional. The forecast for 2026 is broad adoption of navigation alternatives — radio beacons, visual navigation, sensor fusion, and more onboard logic — not because it is elegant, but because it is survivable.
2) “One drone, one operator” is fading
A core expectation for 2026 is a move away from the “one operator per platform” construct and toward platform-level coordination — including swarm middleware, intent-based control, and human oversight that sets rules rather than steering each vehicle.
3) End-to-end design beats component patchwork
Several voices stress that the market will drift from stitching together many vendors’ parts toward integrated end-to-end solutions where hardware and software are designed as a single system — easier to update, harder to break under EW, and more scalable in field conditions.
4) Resilient satellite connectivity: beyond a single provider
Perhaps the most strategically “state-like” forecast is the maturation of alternative satellite communications. One founder points to a GEO-based solution (UASAT GEO) that has seen battlefield use and argues GEO can be a viable alternative to Starlink in scenarios demanding stable, autonomous operation. For 2026, he forecasts the launch of the architecture and operating model for a national UASAT LEO project — explicitly framed as resilient to EW and independent of foreign control.
What 2026 Is Likely to Look Like: Three Bets, Two Risks, One Big Signal
Three bets that appear well-founded (based on 2025 trajectories):
- Interceptor drones as a mass layer, increasingly “software-defined” — with updates and tactics evolving faster than airframes.
- GNSS-denied autonomy becoming mainstream engineering, not a niche feature — with visual/radio-aided navigation and onboard decision-making pushed into standard designs.
- System-of-systems engineering becoming the premium skill — linking platforms, comms layers, autonomy levels and human control into coherent architectures rather than isolated gadgets.
Two risks that could cap performance even if funding rises:
- Talent bottlenecks, particularly embedded + guidance/navigation + systems integration, which are hard to create quickly and already scarce.
- The standards gap — where procurement and certification lag battlefield iteration, forcing teams to “engineer twice”: once for reality, once for outdated rules.
One big signal to watch:
Whether Ukraine’s deftech ecosystem continues its shift “from garage start-ups to corporate business”, with ecosystems of interoperable products rather than heroic single-point solutions. That maturation — organisational as much as technical — is repeatedly framed as a defining dynamic for 2026.