A strike video is not evidence by itself.
It is a claim with images attached.
This distinction matters every time a military video appears on Telegram, X, YouTube, Facebook or a mirrored channel. A camera can show impact. It can show smoke, fire, damage, speed, direction and sometimes a target. But the useful OSINT question is not simply:
Does the video look real?
The better question is:
What can this video prove, what does it only suggest, and what must be verified somewhere else?
The latest example comes from Ukraine’s energy war.
According to Reuters, Russia has been using small FPV drones controlled through fibre-optic cables to damage high-voltage electricity substations in Ukraine’s Sumy region. The footage of the strikes was posted on Russian social media channels, then verified by the Centre for Information Resilience and confirmed by Reuters.
That is the fact pattern worth studying.
Not because the weapon is visually dramatic. Not because every tactical claim around it should be accepted. But because it shows how modern OSINT often works: public video, platform context, infrastructure analysis, battlefield mapping, expert verification and careful limits.
The reported tactic is simple in concept and difficult in practice.
Ukraine has protected important electrical substations with concrete structures and anti-drone netting. These defences are designed to protect expensive and critical transformer equipment from missiles and larger attack drones.
The Russian workaround, as described in Reuters’ reporting, is to use fibre-optic FPV drones.
Unlike many radio-controlled drones, these systems are guided through a physical fibre-optic cable. That makes them harder to stop with electronic warfare as long as the cable remains intact. Instead of relying on a radio link that can be jammed, the drone follows a physical thread back to the operator.
The reported strike sequence is also important.
In the Reuters account, CIR investigator Joshua Scriven describes a pattern in which a first drone damages or opens the protective netting. A second drone then flies through the new gap and strikes the autotransformer.
This is not just a story about drones.
It is a story about adaptation.
One defence changes the target environment. The attacker studies the defence. The weapon changes. The video becomes a public signal that the tactic is being tested, repeated or advertised.
For an OSINT analyst, the video is only one layer.
When a strike video appears on a Russian social channel, there are at least five separate questions.
A real video can still be miscaptioned.
It may show a different date, a different location, a different target or a previous strike reposted as new. The first task is not to admire the footage. It is to detach the visual evidence from the caption.
Useful checks include:
- visual landmarks;
- infrastructure layout;
- road and field patterns;
- shadows and weather;
- visible defensive structures;
- upload history;
- earliest findable version;
- repost chains;
- whether the same clip appears with different claims.
The caption is a lead, not a conclusion.
Energy infrastructure has a recognizable geometry.
Substations often have open yards, transformer bays, access roads, perimeter fencing, gantries and repeated structural patterns. Protective concrete structures and anti-drone netting may create additional visual signatures.
But a recognizable category is not a geolocation.
The analyst still needs to match the video to a specific site. That requires comparing visible features in the footage with satellite imagery, maps, prior imagery, local reporting or other public records.
In this case, Reuters reports that CIR verified multiple strikes, including attacks on large 330 kV substations and smaller 110 kV substations. The locations of the larger strikes were reported to be between 16 and 26 kilometres from the frontline, according to DeepState battlefield mapping cited in the Reuters report.
That distance matters because it helps evaluate a second claim: range.
If verified, the videos do not only show damage. They suggest that small fibre-optic FPV drones are being used at distances that matter operationally.
This is where many social media interpretations overreach.
A clip of a strike may prove that something hit something. It may not prove the full method.
To support the fibre-optic claim, an investigator has to look for additional indicators:
- visible cable or launch evidence;
- repeated strike behaviour;
- known characteristics of similar drones;
- technical assessment from specialists;
- corroboration from recovered components;
- consistency with battlefield reports;
- credible expert review.
Reuters’ reporting relies on CIR verification and additional confirmation. That gives the public reader a stronger basis than an anonymous Telegram caption.
Still, the method should be described carefully:
Public reporting indicates that fibre-optic FPV drones are being used.
That is stronger than:
This specific video alone proves the entire system.
OSINT is not only about location.
A verified strike can answer a narrow question: where did this happen?
A better investigation asks what the event means within a system.
In this case, the target is not a random object. A high-voltage substation is part of a regional electrical network. Reuters cites Ukrainian energy expert Oleksandr Kharchenko saying that disabling the autotransformer in a 330 kV substation can bring down the entire transformer unit and that such a component is worth roughly $3.5 million.
The public-interest question becomes:
Is this a tactical strike, or part of an effort to isolate a region from the national grid?
Reuters reports that CIR interprets the strikes as part of a wider strategy to reduce the ability of a region to receive electricity from outside and then degrade local generation.
That is an analytical assessment, not a video fact.
It should be labelled as such.
For ProjectOSINT, this case is useful because it gives us a compact workflow for strike videos.
Save the post, the video, the caption, the account name, the timestamp, the URL and any visible engagement/context.
If the platform is Telegram, preserve the channel post link when possible. If the clip is reposted elsewhere, keep the repost chain but do not treat the loudest post as the earliest source.
The first question is:
What did the public record look like when we found it?
Do not verify the whole caption at once.
Break it down.
- This is a fibre-optic FPV drone.
- The target is a Ukrainian substation.
- The substation is in Sumy region.
- The strike happened recently.
- The drone bypassed protective netting.
- The target was an autotransformer.
- The tactic is being used repeatedly.
- The attacks are part of a regional blackout strategy.
Each claim needs a different kind of evidence.
Some may be visible in the video. Some may require maps. Some may require expert analysis. Some may remain hypotheses.
Explosions are visually noisy.
Infrastructure is usually more useful.
Look for the stable features around the strike:
- transformer yard layout;
- access roads;
- fence lines;
- nearby tree lines;
- pylons and power lines;
- concrete protective structures;
- netting frames;
- rail or road access;
- building shadows;
- distinctive roof or ground patterns.
The goal is not to identify a general type of place. It is to connect the footage to a specific physical site.
A video can be geolocated and still be old.
Chronolocation may use:
- first upload time;
- weather;
- vegetation;
- shadow direction;
- visible damage before and after;
- local outage reports;
- official statements;
- satellite imagery;
- other videos from the same incident.
If the timing cannot be confirmed, say so.
Every strong OSINT piece should contain a limit statement.
For this case:
- A public video may support the fact of a strike.
- Geolocation may support the target location.
- Repeated verified videos may support a pattern.
- Expert analysis may support the fibre-optic method.
- But strategic intent requires more caution.
The sentence should be boring and useful:
The available public evidence supports the existence of a repeated strike pattern against substations, but it does not by itself prove the full command intent behind the campaign.
That kind of sentence protects the investigation from becoming propaganda.
The visible story is about drones.
The OSINT story is about evidence discipline.
Fibre-optic FPV drones are important because they show how quickly battlefield methods adapt to defensive measures. Concrete protection, anti-drone nets and electronic warfare do not end the problem. They change the attack surface.
But the public information environment adapts too.
Military actors publish videos because videos persuade. They compress a complex event into a few seconds of apparent certainty. The viewer sees a drone approach, a gap, a target, an impact. The brain wants to complete the story.
OSINT has to slow that process down.
Not because the footage is useless.
Because the footage is only the beginning.
The method is:
preserve, separate, geolocate, corroborate, classify, limit.
If a strike video survives that process, it becomes evidence.
If it does not, it remains a claim with images attached.
That difference is the work.
Russia’s unstoppable fiber-optic drones shutting down Ukraine’s power grid piece by piece
NATO-aligned media is sounding the alarm about the capabilities of Russia’s unjammable small, cheap drone fleets, reporting eight separate, successful attacks on 110 and 330 kV substations (four of each type) in Sumy, Ukraine, 16-26 km from the frontlines.
🌏 Reuters describes the drones’ complex, tandem-charge and follow-on strike tactics, involving one UAV breaking through anti-drone netting before a second flies through.
🌏 The second drone then maneuvers around concrete sarcophagi structures Ukraine has built to try to protect its grid, straight into ventilation holes, to knock out substations that cost up to $3.5M apiece.
🌏The price of the drones? $2k. Immune to Ukraine’s NATO-provided electronic warfare systems, the small fiber optic FPV strike drones make short work of on-site defenses.
🌏 The news agency did not go into details on the types of drones involved, but we know Russia’s FPV fiber-optic drone fleet includes:
♦️ Knyaz Vandal Novgorodsky – 30km range, 3 kg payload. Among the most widely available and used UAVs of this kind in Russia’s arsenal.
♦️ Molniya-2: Ultra-cheap, fixed-wing drone made of plywood, polystyrene and aluminium tubing. 6-10kg payload, 40-50km range, production cost as low as $500.
♦️ Skvorets: another FPV fiber-optic-ready platform. 3.5 kg payload, 8km range. 150km/h top speed.
♦️ Veterok: 7 kg payload, 30 km range. Designed for minelaying and targeting of armored vehicles.
The Knyaz Vandal Novgorodsky pioneered FPV fiber-optic drone warfare, debuting in August 2024 during the campaign to expel Ukrainian forces from Kursk region. It was developed by SPC Ushkuynik, a civilian-volunteer-backed tech group based in Veliky Novgorod.
Created to counter the proliferation of vehicle-mounted and stationary radio jammers which degrade FPV signals, the ingenious concept is a nod to the 20th century wire-guided tech used in anti-tank guided missiles.
If this is useful, share it.
This is the weekly selection. But it’s not the only one.
If you’d like to read more: → full articles on the website
If you’d like to get the latest updates first: → Telegram
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.