Weekly review, week to 1 July 2026

Key structural shifts

1. The crisis of classical air defense in the face of the "cheap drone versus expensive missile" asymmetry.

The defining process of the week is the transition of Ukrainian strikes on Russia's rear from the category of "episodic pinpricks" into the category of systemic strangulation. The quantitative milestone: growth in long-range UAVs from >5000/month in spring 2026 to ~10,000/month by June; the largest recorded raid — 660 drones in one night across 12 regions. The key thesis, now acknowledged by the Russian side as well (@milinfolive, @chadayevru): the post-Soviet air defense architecture, built for classical threats, is structurally incapable of countering massed attacks. There is "nowhere to get" additional SAM systems — the only answer is acknowledged to be "fight fire with fire" (cheap interceptors + turrets). This is a fundamental, not a situational, shift.

1. The logistics (fuel) war as a theater of its own.

Strikes by "Gerans" and Hornet loitering munitions have shifted from one-off "video clips" to methodical work against refineries, tank farms, fuel and lubricant depots and unprotected road logistics (~800 strikes on trucks since May). The flip side is a fuel crisis in Russia (gasoline rationing by QR code/license number in a number of regions of the Southern Federal District, a cut in truck mileage from 700→500 km, −9% in ammonium nitrate production). An important analytical nuance (@chadayevru): the strikes did not create the crisis but exposed years of accumulated underinvestment in oil refining — that is, kinetics became a catalyst for a structural vulnerability of the rear.

1. Ukraine's entry into the club of long-range high-end systems.

The visually confirmed use of the FP-5 "Flamingo" cruise missile (a strike on "Titan-Barrikady" in Volgograd, at least 2 of 5 got through air defense), a probable interception of a Ukrainian ballistic missile over the Moscow region, systematic strikes on 4 space communication centers ("Dubna", "Vladimir", "Medvezhyi Ozera", "Skolkovo"). Plus negotiations with France on a SCALP license and a contract for 32 Gripen E/C/D. A mature multi-layered long-range arsenal is taking shape: cruise missiles + ballistic missiles + mass drones.

Technology dynamics and trends

Rising vector — autonomous AI terminal guidance as the norm (scaling).

This is the main qualitative transition of the week: from experiments to serial production.
- [TECH] The "Geran-Seeker" is an option for all versions of the "Geran" (2–4), more often on the jet-powered "Geran-4": several cameras, a MESH radio modem (a cellular self-organizing radio network) and, critically, an automatic target recognition/lock-on system with automatic guidance based on a Raspberry Pi. The direct consequence: low-power EW (electronic warfare) on the target no longer saves it — machine vision brings the munition to the target without an operator and without a radio link in the terminal phase.
- [TECH] A matching signal: the generation of a full 3D model of the "Zaliv" shipyard in Kerch in 30 minutes from video and satellite imagery — a transition from flat images to volumetric models for training AI in munitions. Two independent signals add up to a single chain: target dataset → trained machine vision → autonomous terminal lock-on.

Rising vector — the formation of an interceptor drone class.

Over the week the segment took final shape, with a price "anchor" of ~$2000:
- [TECH] ZIRKA (Ukraine, NOCTIS/Vyriy): >340 km/h, ceiling 6 km, up to 20 min, auto-detection + terminal guidance, up to $2000; "Talion" (Ukraine).
- [TECH] "Molniya-P" (Russia): MSL-20045 compact radar, detection of FPVs at up to 8 km and of large targets at up to 15 km; "Rita" has been announced.
This is a transition from improvised FPV rammers to serial products with their own radar and automatic guidance.

Rising vector — robotization of the "last mile" and reduction of personnel losses.

[TECH-lite] Ukrainian practice: moving FPV crews and supplies forward on UGVs (unmanned ground robotic systems) closer to the front line, launching 5–6 drones from a single ground platform in "ambush" mode. The logic — "lose electronics, not people". The "Rubikon" dataset (35,000+ targets) confirms the mirror-image priority: uncrewed systems (UAVs+UGVs) account for 39.6% of hits (and growing), plus a targeted hunt for the "nervous system" — Starlink terminals, patch antennas, EW, communication nodes (15.4%). The pace has accelerated radically: the interval between 5000-target "milestones" has shrunk from 297 to 30 days.

Degrading vector — the effectiveness of classical countermeasures.

In Russia, mobile fire groups, anti-aircraft drones and pontoon bypasses have been acknowledged as ineffective against massed attacks; a symmetrical long-range response is limited (only guided "Shaheds" are reliable, but they do not compete on price/pace). Rear-area C-UAS is being institutionalized through numbers (8 BARS-Belgorod battalions, 7 on C-UAS) but remains technologically low-tech (improvised turrets for twin 7.62-mm PKT machine guns — 54 kills/day). The development vector is stated explicitly: "smart" turrets + HPM (High Power Microwave, high-power microwave weapons).

EW resistance as a separate track.

Satellite communications (Starlink) are becoming the standard for supplying/controlling infiltrating groups; [TECH] "Rosel"/Rostec demonstrated communications with frequency hopping (adapting to interference); the ZALA Z-16 with onboard AI evades interceptors and scans their frequencies. EW as the sole countermeasure is becoming a thing of the past from two directions at once — through machine vision (bypassing jamming) and through frequency adaptation.

Weak signals (Under the radar)

• Battlefield electrification as a logistical ceiling. [TECH] A translation of a RUSI report (@kcpn2014): Li-ion batteries do not work in the cold (confirmed by the current conflict), and energy is becoming a scarce logistical resource on a par with fuel and ammunition. This is a fundamental constraint on all "last mile" robotization and on interceptors — almost unnoticed amid high-profile videos. For the engineering community: battery thermal management and alternative chemistries are becoming a critical node.
• Laser wireless in-flight charging of UAVs (SolaNika, Japan, first stage ~1 km). A direct answer to the energy ceiling above — if it scales, it removes the loiter-time constraint for interceptors and reconnaissance drones.
• The systemic sham of Russian drone localization. [TECH] An insider analysis of the Ministry of Industry and Trade's points system (@UAVDEV): "Baikal-U flight controllers simply don't turn on", an STM32H743 does the actual work, while a Russian chip blinks next to it "for show". The strategic consequence: Russia's declared localization is illusory, and dependence on the Chinese component base persists — a vulnerability under sanctions pressure.
• China's containerized modular weapons (@china3army): up to 2000 sets/year, rapid conversion of civilian vessels into military carriers — a marker of the future architecture of naval warfare, underestimated against the background of the Ukrainian agenda.

Vector projection and implications

Confirmed dynamics (high confidence):

• Terminal AI guidance devalues point EW. The "Geran-Seeker" on a Raspberry Pi + 30-minute generation of 3D datasets mean that jamming at the target itself loses its point. The required countermeasures are shifting: from suppressing the radio link to [TECH] optical/laser dazzling of sensors, signature masking (thermal/visual), smoke and physical shielding of critical nodes (tank farms, power lines, space communication centers). EW retains value only against non-autonomous drones in the cruise phase.
• The race of $2000 interceptors will continue to squeeze prices and increase detection autonomy. The next point of development is a unified "detection → target allocation → interceptor swarm" loop ("Tsitadel"/"Zubr" turrets + interceptor drones in one network). The British CCV (Common Combat Vessel) and Helsing's German CFSN combat cloud (€580 million) follow the same principle at the fleet/air force level: a crewed "conductor" + a swarm of autonomous platforms.
• Fuel-and-logistics pressure on Russia will intensify. There is as yet no technical countermeasure against massed cheap drones on area targets (refineries burn in "controlled burnout" due to a lack of firefighting equipment). The expected response is dispersal of fuel and lubricant storage, physical protection of tanks, and emergency deployment of HPM systems over processing nodes.

Analytical hypotheses (moderate/low confidence):

• The energy ceiling (RUSI: Li-ion in the cold) + laser charging (SolaNika) point to the next node of competition — onboard power supply. Hypothesis: by the fall-winter season, the side that has solved thermal management/recharging will gain an advantage in the loiter time of interceptors and reconnaissance.
• The Pentagon's shift in targeting doctrine toward an "AI initiates — human controls" model + the Ukrainian DOT-Chain/"Bazovyi Riven" platform (guaranteed drone supply with automatic selection of an equivalent) point to the algorithmization of the entire kill cycle. Hypothesis: the software layer (combat clouds, automatic targeting) will become the decisive differentiator sooner than hardware.
• The sham of Russian localization (STM32H743 instead of Baikal-U), combined with the accelerating pace of "Rubikon" hits, gives rise to a component bottleneck hypothesis: if control over Chinese re-exports of flight controllers is tightened, the Russian drone industry risks hitting a serial-production ceiling sooner than the Ukrainian one, which relies on the European relocation of production (Latvia, PGZ, exchange of drone know-how for equipment).