Emergency drones change course without warning. Delivery drones follow the same path every time.
Above the suburbs of North Texas last spring, NASA orchestrated a quiet but consequential experiment: teaching the sky to make way for urgency. As unmanned aircraft multiply over American cities — delivering parcels, inspecting bridges, monitoring crowds — the question of who yields to whom in shared airspace has moved from theoretical to pressing. Working alongside the FAA, local police departments, and private technology partners, NASA demonstrated that the protocols governing emergency drones can be built to mirror something humanity already understands: the way traffic parts for a siren.
- Urban skies are filling faster than the rules governing them — commercial drones now share airspace with emergency responders who cannot file a flight plan mid-pursuit.
- The core tension is behavioral: delivery drones follow predictable corridors, while police and rescue drones lurch and pivot in real time, creating cascading disruptions for every aircraft nearby.
- NASA simulated an aerial vehicle pursuit — a drone shadowing an erratically driven SUV — to measure exactly how sudden, unplanned course changes ripple through surrounding commercial traffic.
- Automated systems, not human operators alone, proved necessary to respond quickly enough: when a public safety drone launched, nearby commercial aircraft received real-time instructions to yield, much like cars pulling over for an ambulance.
- The North Texas exercise confirmed the concept is viable; the harder work ahead is scaling these protocols into the permanent operational infrastructure of American airspace.
Last spring, the skies above North Texas became a laboratory. NASA launched emergency response drones into airspace already occupied by commercial delivery and inspection aircraft, asking a deceptively simple question: when a police, fire, or rescue drone needs to move fast and without warning, how does the system around it respond?
NASA's Air Traffic Management and Safety project, based at Ames Research Center, has spent years building toward an answer. The North Texas exercise brought together the FAA, the Texas Department of Public Safety, and police departments from Fort Worth, Arlington, and Irving, alongside industry partners whose airspace management platforms provided the technological backbone. When a public safety drone entered the simulation, surrounding commercial drones received instructions to yield — the aerial equivalent of traffic parting for an ambulance.
What the exercise clarified most sharply was how different emergency drone behavior is from commercial operations. A delivery drone flies a pre-approved corridor with predictable efficiency. A first responder drone chases a suspect, tracks a spreading fire, or searches for a missing person — reacting to the ground below in ways that make advance route planning impossible. To stress-test this unpredictability, NASA had a drone follow an officer driving an SUV erratically, generating data on how sudden directional changes ripple through surrounding traffic.
The findings pointed toward a system that must be faster than human reaction alone — automated, communicative, and adaptive in real time. Project manager Shivanjli Sharma framed the exercise as a foundation: building the data, tools, and frameworks needed to make future drone operations safe and scalable for commercial operators and emergency responders alike. The concept has been proven. The work now is to make it permanent.
The sky above North Texas filled with drones last spring, but not the kind most people imagine. Alongside the commercial delivery aircraft and inspection vehicles that increasingly populate urban airspace, researchers from NASA launched something more urgent: emergency response drones that needed to move through crowded skies without warning, changing course suddenly, responding to situations unfolding on the ground below. The question driving the exercise was straightforward but operationally complex: when a police, fire, or rescue drone needs to move through shared airspace quickly, how does the system respond?
NASA's Air Traffic Management and Safety project, headquartered at Ames Research Center in Silicon Valley, has spent years grappling with this problem. As cities fill with unmanned aircraft—some delivering packages, others inspecting infrastructure, still others conducting surveillance—the challenge of keeping them safely separated has grown urgent. The FAA oversees all civil airspace in the United States, but the agency needed tools and protocols that didn't yet exist. The North Texas exercise, conducted in partnership with the FAA, the Texas Department of Public Safety, and police departments in Fort Worth, Arlington, and Irving, tested whether those tools could actually work.
The demonstration brought together industry partners like Drone Sense, Avision, and ANRA Technologies, whose airspace management platforms formed the technological foundation of the test. When a public safety drone launched during the simulation, nearby commercial drones received instructions to move aside—much like traffic yielding to an ambulance with sirens blaring. When multiple emergency agencies responded simultaneously to simulated incidents, their operators coordinated in real time to determine which drone operations should receive priority at any given moment. Abhay Borade, a research lead on the project, explained the underlying philosophy: the goal was not to benefit emergency responders at the expense of commercial operators, but to find approaches that prioritized safety for everyone while ensuring efficient use of the shared airspace.
What emerged from the exercise was a clearer understanding of how fundamentally different public safety drone operations are from commercial ones. A delivery drone typically follows a pre-planned route, logged and approved in advance, flying the same corridor from warehouse to doorstep with predictable efficiency. A first responder drone operates in an entirely different mode. During a search-and-rescue mission, a vehicle pursuit, or when scanning a hazardous environment, these aircraft must change course suddenly and without warning. They respond to real-world developments unfolding below—a suspect changing direction, a fire spreading unexpectedly, a missing person spotted in a new location. Advance route planning is essentially impossible.
To understand how these unpredictable movements might affect nearby commercial traffic, NASA researchers designed a creative test: a drone tasked with following an officer driving an SUV erratically, mimicking an aerial pursuit of a fleeing vehicle. This exercise generated valuable data on how sudden, unplanned directional changes by a public safety drone ripple through surrounding airspace, forcing nearby commercial drones to respond and adapt in real time. The insights proved critical. They showed not just that the system could work, but how it needed to work—with automated tools that could respond faster than human operators alone, and with communication protocols that kept all parties informed.
Shivanjli Sharma, the project manager for Air Traffic Management and Safety at NASA Ames, described the exercise as a step toward building the data, tools, and traffic management frameworks needed to ensure future drone operations are safe, responsible, and scalable for everyone. As commercial drone activity continues to grow and public safety agencies increasingly rely on unmanned systems for law enforcement, firefighting, and emergency response, the need for sophisticated, adaptive airspace management will only intensify. NASA's collaborative approach—bringing together federal regulators, local agencies, and private industry—positions the agency to lead this effort. The North Texas exercise demonstrated that the concept works. What comes next is scaling it, refining it, and embedding it into the operational fabric of American airspace.
Notable Quotes
The goal is not simply to benefit emergency responders at the expense of commercial operators, but to find approaches that prioritize safety for all while ensuring efficient use of shared airspace.— Abhay Borade, research lead, NASA Air Traffic Management and Safety project
We're helping build the data, tools, and traffic management frameworks needed to ensure the future of drone operations is safe, responsible, and scalable for everyone.— Shivanjli Sharma, Air Traffic Management and Safety project manager, NASA Ames
The Hearth Conversation Another angle on the story
Why does a delivery drone need to move differently than an emergency drone? Can't they just follow the same rules?
A delivery drone knows its route before it takes off. It's optimized for efficiency—same path every time, logged with the FAA. An emergency drone doesn't have that luxury. A police officer pursuing a suspect, a firefighter scanning a burning building, a rescue team searching for a missing person—they can't plan their route in advance. They respond to what's happening on the ground in real time.
So the system has to be reactive, not predictive.
Exactly. And that's the hard part. Commercial drones are predictable. You can plan around them. But when a public safety drone suddenly banks left to follow a fleeing vehicle, every other drone in that airspace has to know about it instantly and get out of the way. That's not something a human operator can manage alone.
How did they test whether it would actually work?
They had a drone follow an SUV driving erratically through North Texas, simulating a vehicle pursuit. They watched how the system responded—how commercial drones nearby reacted, how quickly they could move aside, what happened when multiple emergency agencies needed priority at the same time.
And it worked?
It worked well enough to show the concept is sound. But there's a difference between a test and reality. In a real emergency, with lives at stake, the margins for error shrink. That's why they're gathering this data now—to make sure the system is robust before it's actually needed.
What happens if two emergency drones need the same airspace at the same time?
That's where the coordination comes in. The operators talk to each other in real time, decide which response is more urgent, which drone gets priority. It's like dispatch coordinating multiple ambulances to the same area—someone has to make the call about who goes where.