EyesOn · 2026-04-15

What Actually Happens to Your Drone Video Feed When You Use a SaaS Platform

You launch your DJI Matrice 30T, connect to whatever cloud streaming service your organization signed up for, and your video is live. Incident command can see it. Your supervisor can see it. Maybe a few other stakeholders can see it.

Who else can see it?

That question doesn't get asked enough. And when you start tracing where drone video actually travels inside a SaaS streaming pipeline, the answer gets uncomfortable fast — especially if your missions involve sensitive locations, law enforcement coordination, private property, or vulnerable individuals.

This isn't a hypothetical concern dressed up as marketing. It's an operational security issue that every professional drone operator running live video should understand before they fly another mission.

Where Your Video Feed Actually Goes

When you stream through a managed SaaS drone platform, your video travels a specific path — and that path involves infrastructure you don't control, terms of service you probably didn't read carefully, and data retention policies that may store your footage long after you think it's gone.

The typical SaaS pipeline

Your drone captures video. Your controller encodes it. That encoded stream goes out over the internet — usually to a cloud relay server operated by the SaaS company. From there it gets distributed to your viewers. Latency is introduced at each hop. A round trip from your drone through a commercial cloud relay and back to a viewer who is standing fifteen feet from you can take two to eight seconds. That's not a quirk — it's structural. The data has to travel to a server farm somewhere in Virginia or Oregon before it comes back.

But the latency is almost beside the point when you're thinking about security. The real issue is that your video feed exists, in clear form, on someone else's hardware, processed by someone else's software, under someone else's terms.

What those terms usually say

Most drone streaming SaaS agreements include some version of the following clauses:

That last one matters more than operators typically acknowledge. If your drone video passes through a commercial SaaS provider's servers, that provider can receive a subpoena, a National Security Letter, or a law enforcement request — and respond to it without notifying you, depending on jurisdiction and legal circumstance. Your footage, your GPS track, your flight timing — all of it becomes discoverable through a third party you don't control.

For a real estate photographer, this is probably fine. For a security contractor surveilling a client's industrial property, a search and rescue team operating in coordination with law enforcement, or a public safety agency running sensitive operations — the implications are materially different.

The Self-Hosted Difference — What It Actually Means in Practice

Self-hosting isn't a philosophical stance. It's an operational decision with concrete security consequences.

EyesOn runs on your server. Your hardware. Your network. The video stream from your drone controller goes directly to your infrastructure — not through a cloud relay in a data center you've never visited, operated by a company you've never audited. When your stream is live, it exists on machines you control. When the mission is over, what happens to that footage is entirely your decision.

There is no third party positioned between your operator and your incident command who can receive a legal demand for your data, experience a breach, get acquired, change their terms of service, or go out of business.

What EyesOn's architecture actually does

EyesOn uses WebRTC for peer-to-peer video delivery. The stream originates from the companion Android app, which captures the full DJI controller screen including OSD data — airspeed, altitude, GPS coordinates, battery status, everything your operator sees. That full-context stream delivers to authorized viewers at approximately 200ms latency. Not 2-8 seconds. 200 milliseconds.

That's not a cloud relay inserting itself into the path. That's a direct WebRTC connection through infrastructure you own.

The server runs in Docker. You provision it. You control the SSL certificates. You control who has viewer credentials. You control the logs. You control retention. If you want footage stored, you store it. If you don't, you don't. No one else has a copy.

Local AI-assisted reporting runs via Ollama and Mistral 7B directly on the server. Flight analysis, reporting, mission documentation — none of that data leaves your network. It doesn't get sent to an API endpoint somewhere for processing. The computation happens locally.

Real Operational Scenarios Where This Matters

I want to be specific about what self-hosted video security means in the field, because the abstract argument only goes so far.

Law enforcement coordination

I've flown SAR missions coordinating directly with Lane County Sheriff's Office. LCSO shared cell ping coordinates with me — last known position data for a missing person. That kind of operational data, combined with live aerial video, is exactly the kind of information that needs to stay inside a controlled information environment. It doesn't belong on a SaaS platform's servers. It belongs on controlled infrastructure with a clear chain of custody.

When your drone video is routed through a commercial third party, you've introduced a link in that chain that you don't control. For routine commercial work, that's acceptable risk. For anything touching law enforcement data, vulnerable individuals, or sensitive investigative operations, it's not.

Industrial security surveillance

I've run overnight patrol laps over industrial yards using the M30T's thermal sensor. When the thermal picked up a heat signature near a trailer during a scheduled 2:30 AM patrol, that footage — the detection, the identification, the incident — is evidence. It has chain-of-custody implications. It may be used in legal proceedings. The question of where that video resided, who could have accessed it, and what the retention policy was becomes relevant the moment a suspect's lawyer starts asking questions.

Self-hosted means I can answer those questions clearly. The footage was on the client's server. Access logs exist. No third party had access. The chain is clean.

Sensitive geographic data

Thermal sweeps over private property generate data that property owners have a reasonable expectation will stay private. Grid-pattern flight coverage over 800 acres generates detailed information about terrain, structures, and features that the landowner didn't consent to share with a cloud provider's infrastructure team. When you stream that footage through SaaS, you're making decisions about third-party data exposure that aren't entirely yours to make.

Pricing Transparency: The Security Cost of SaaS

SaaS drone streaming platforms don't just have privacy implications — they have significant cost structures that operators often don't fully account for.

DroneSense runs $1,500 to $5,000 per year per drone. If you operate three aircraft, you're looking at $4,500 to $15,000 annually. FlytBase meters by viewer minute, which means your cost scales with the size of your incident command. A major operation with twenty viewers watching a four-hour flight can generate a billing surprise you weren't expecting.

EyesOn Personal is $149 setup and $39 per month — $617 for the first full year, with no per-drone limit and no per-viewer limit. Your fleet can grow. Your audience can grow. Your bill doesn't change.

Professional is $299 setup and $89 per month, supporting up to five server deployments with email support. Enterprise is $499 setup and $209 per month — unlimited server deployments, priority support, custom branding. Managed hosting by BarnardHQ with a dedicated support SLA runs $799 setup and $499 per month for operators who want the self-hosted security model without managing the infrastructure themselves.

None of those tiers route your video through BarnardHQ's servers. Even Managed hosting keeps your data environment isolated. The business model is software and support — not data.

What Happens When You Don't Renew

This is a scenario worth thinking about before you need to think about it.

With every SaaS drone streaming platform I'm aware of, a lapsed subscription means a locked-out account. Your historical data may be held, deleted, or made inaccessible. Your workflows stop working. You're back to zero until you renew or find a replacement.

EyesOn keeps running. The software doesn't phone home to validate your subscription. If your subscription lapses, you stop receiving updates and support. The server you deployed keeps streaming. Your operational continuity doesn't depend on a billing relationship with a company in a different city that you've never met.

That's not a quirk of the business model. It's a deliberate decision about what a tool you depend on should behave like.

The Operational Security Baseline

If you're running drone operations where the content of your video feed has any sensitivity — law enforcement coordination, private property surveillance, SAR missions involving vulnerable individuals, industrial security, government contracts — the baseline question is not which SaaS platform has the best interface. The baseline question is: who can see this video besides the people I've authorized, and under what circumstances?

With self-hosted EyesOn, the answer is clear. Your server. Your credentials. Your footage. Your rules.

That answer is harder to give with a SaaS platform — not because those companies are necessarily bad actors, but because the moment your data lives on their infrastructure, the question isn't fully yours to answer anymore.

For operators who need to be able to answer it clearly: eyeson.barnardhq.com.

DroneStreamingOperationalSecuritySelfHostedEyesOnDroneOpsDataPrivacyWebRTCPublicSafetyDroneVideoBarnardHQDJIM30TSARDronesCyberSecurityDroneOperatorPart107
← Back to all posts