Seventy-seven minutes in Uvalde. These drones arrive in fifteen seconds.
In the space between an alarm and a police siren, children have long been left unprotected — and it is into that silence that a new kind of guardian has been proposed. Campus Guardian Angel, an American company drawing lessons from drone warfare in Ukraine, is testing remotely piloted drones designed to reach an active school shooter within fifteen seconds of an alarm. Pilot programs are underway in Florida and Georgia, funded with public money, as the United States confronts a crisis that claimed lives in 233 school shooting incidents in 2025 alone. The question being asked is not merely technological, but moral: whether a society that has failed to prevent this violence can at least shorten the time its children are left alone with it.
- With 233 school shooting incidents recorded in 2025 and the memory of Uvalde's 77-minute police response still raw, the urgency to close the gap between alarm and intervention has never felt more desperate.
- Campus Guardian Angel is deploying sub-kilogram drones capable of 65km/h, launched from hidden hangars inside school buildings and controlled remotely from Austin, Texas — a system adapted directly from battlefield drone tactics observed in Ukraine.
- Every decision to strike, warn, or deploy pepper gel rests with a human pilot — not an algorithm — a deliberate design choice meant to address fears of autonomous lethal force in school hallways.
- Pilot programs in Florida and Georgia are already running on public funds, and Houston parents have signaled willingness to pay out of pocket, suggesting demand is outpacing the debate.
- The system's expansion now hinges on whether school districts will absorb the cost and whether communities are prepared to accept armed drones as the first face of protection their children see in a crisis.
A teacher triggers an alarm. Police are minutes away. In that interval, Campus Guardian Angel believes a squadron of small drones can already have reached the threat.
The company is testing exactly this system in American schools right now, drawing its inspiration from an unlikely source: the drone warfare reshaping battlefields in Ukraine. Watching how fast and effective first-person-view drones had become in Eastern Europe, the company's leadership asked whether that same speed could protect schoolchildren. The answer became a product.
Each school is mapped in three dimensions. Small hangars — installed in hallways and around perimeters — each hold three drones weighing under a kilogram, spanning roughly twenty-five centimeters, and capable of reaching 65km/h. When an alarm sounds, human pilots at a command center in Austin take control, with a target response time of fifteen seconds. The pilots are not soldiers or security veterans — many are competitive drone racers, people fluent in the language of controllers and screens.
What the drones do upon arrival depends on the situation. A minor brandishing a weapon but not yet firing might be met with verbal commands through two-way audio. An active attack triggers a harder choice: kinetic impact or non-lethal pepper gel. Every call is made by a person, not an algorithm — a line the company has drawn deliberately.
The scale of the problem gives the project its weight. In 2025, there were 233 firearm incidents on American school campuses. In Uvalde in 2022, nineteen students and two teachers were killed while police took seventy-seven minutes to act. That gap is what this system is designed to collapse.
Pilot programs are running in Florida and Georgia with public funding. In Houston, parents have expressed readiness to pay themselves. The company's co-founder, a former Navy SEAL, describes the ideal outcome simply: install the system everywhere and never need it. Whether that vision scales depends on cost, community trust, and a question America has not yet answered — what it is willing to place between its children and the next alarm.
A gunman enters a school. A teacher hits an alarm on their phone. Police are minutes away. But before they arrive, a squadron of small drones descends from hidden stations and stops the shooter in seconds.
This is not science fiction. Campus Guardian Angel, a U.S. company, is testing exactly this system right now. The drones are real, they are being deployed in pilot programs, and they represent a direct response to one of America's most persistent crises.
The idea came from watching Ukraine. Khristof Oborski, the company's director of tactical operations, explained that the concept emerged when the firm's leadership observed how effective first-person-view drones had become on the battlefield in Eastern Europe—fast, hard to stop, and capable of reaching targets before traditional forces could. The question became obvious: what if that same speed and precision could protect American schoolchildren?
The system works like this. Campus Guardian Angel maps each school in three dimensions, plotting routes through hallways and classrooms. Then they install small hangars in strategic locations—inside buildings, outside perimeters—each containing three drones. When an alarm sounds, human pilots sitting at a command center in Austin, Texas, take control. The goal is to reach the threat in the first fifteen seconds. The drones themselves are small: less than a kilogram, about twenty-five centimeters across, capable of speeds up to sixty-five kilometers per hour. They are manufactured in the United States.
What happens next depends on what the shooter is doing. If it's a minor carrying a weapon but not yet firing, the drones—equipped with two-way audio—can issue commands and warnings. If children are actively being attacked, the company has prepared two options: kinetic impacts, meaning the drones strike the person, or deployment of non-lethal pepper gel. Every decision is made by a human operator, not by artificial intelligence. Oborski emphasized this point. The pilots themselves are not soldiers. They are competitive drone racers, people more comfortable with video game controllers than weapons. One pilot, Alex Campbell, described the work as gratifying—knowing that he might help officers go home safe and ensure children do the same.
The scale of the problem these drones are meant to address is staggering. In 2025 alone, there were 233 incidents involving firearms on school campuses across the United States. In May 2022, a gunman entered Robb Elementary School in Uvalde, Texas, and killed nineteen students and two teachers. The police response took seventy-seven minutes. That gap—between the moment violence begins and the moment trained responders arrive—is what Campus Guardian Angel is trying to collapse.
Pilot programs are already underway in schools in Florida and Georgia, funded with public money. In Houston, parents have expressed willingness to cover the costs themselves. The drones are offered through annual contracts, with pricing that scales based on school size and the number of buildings. Bill King, a former Navy SEAL and company co-founder, articulated the ideal scenario: install the system in every American school and never have to use it. But that vision depends on whether schools and districts will accept the cost, and whether communities will embrace armed drones as a first line of defense. The technology exists. The question now is whether America will deploy it.
Notable Quotes
The ideal scenario would be to install this system in every single school in the United States and never have to use it.— Bill King, former Navy SEAL and co-founder of Campus Guardian Angel
It's gratifying to know you can help officers do their job, go home safely, and make sure all these children get home safely too.— Alex Campbell, drone pilot
The Hearth Conversation Another angle on the story
How do you decide when a drone should use force versus just issue a warning?
The operator in Austin makes that call in real time, watching what's happening through the drone's camera. If someone's just carrying a weapon but hasn't fired, you talk them down. If they're actively shooting children, you don't hesitate.
That's a lot of responsibility for one person in a control room.
It is. But it's also faster than waiting for police to breach a building. Seventy-seven minutes in Uvalde. These drones can be there in fifteen seconds.
Why not use artificial intelligence to make those decisions automatically?
Because a machine can't understand context the way a human can. A kid with a gun looks different from an active shooter. An AI might not see that difference. Humans do.
The drones came from watching Ukraine. Does that bother you—that school safety is now modeled on warfare?
The technology came from there, yes. But the intent is completely different. In Ukraine, drones are weapons of war. Here, they're meant to stop violence before it spreads. It's the same tool, different purpose.
What if the operator makes a mistake?
That's the risk you take with any armed response. But the alternative is waiting. And waiting has a cost too.