By Fritz Reber, Captain (Ret.), Chula Vista Police Department
People sometimes call me the Father of DFR, or the Godfather of DFR. I'm happy with either one, and I'll own where the ideas came from. The Live911 concept was 100 percent mine. Drone as First Responder grew out of SARA 1, the program I built with Lieutenant Eric Thunberg, and I conceived DFR itself in collaboration with Kabe Termes. But having an idea is one thing, and making it work is another. So many ideas hit a wall because there aren't skilled people to carry them out. DFR worked because a progressive chief, a federal pilot program, and a lot of talented people took those ideas and put them into action, and they deserve a big share of the credit for its success. The Origin Story page on this site gives you the short version. This is the long one, start to finish, as best I can tell it.
My full name is William Reber. I carry both of my grandfathers' names, William and Frederick, and Fritz was a nickname that came from one of them. It stuck from the earliest days. At first I hated it. Now I figure it's memorable, and that's not so bad.
I spent 27 years with the Chula Vista Police Department, in San Diego County, about eight miles north of the Mexican border. Over that time I worked a little bit of everything. I was on SWAT for about five years. I worked undercover and on a task force. I worked investigations and later managed the investigations units, including homicide and narcotics. I did a stint in internal affairs. I ran the dispatch center and I ran the jail. I retired in May 2018 as the patrol operations captain, which was the number two position in the department.
Looking back, I was always the guy poking at technology to see if it could solve an everyday problem. Years before DFR, I worked with a company to develop a GPS tracking device we could put on stolen cars, before that was really a thing. And I was a drone hobbyist long before I had any formal role with drones.
SWAT is where the seed of that interest got planted. As a SWAT operator, one of your biggest concerns is going into a room without knowing what's in there. That stayed with me. By the time I was a lieutenant in investigations, I was buying little first-person-view drones and mini whoops and practicing with them indoors, imagining a SWAT mission. I'd see how far away from a house I could be and still fly in through the front door, and whether the signal would hold through the walls. Then I bought a DJI Phantom 4, and later a Mavic. What blew me away was return to home. At my house in Alpine I'd fly the thing out until I couldn't see it, push a button, and watch it come back and land within a few square feet in my backyard. I'd sit on the couch and listen to it hum as it landed. But that idea of automation, of a machine that could do the hard part on its own, never left me.
As a brand-new lieutenant, not long after the 2008 financial crash, the city was cutting costs, and the department needed one sworn lieutenant to take over both the jail and the communications center. Our Support Operations captain asked for a volunteer. I could see how it would go: everyone else would step back and the new guy would be left standing there. So I walked into his office and volunteered before anyone could volunteer me. I had no idea what I was getting into.
It was the first time I had supervised civilians instead of sworn officers. Running a jail was its own education. Running a 911 center, or PSAP, was another one entirely. I'll admit that when I took over dispatch, I had never even heard the term PSAP. Like a lot of cops, I thought I understood dispatch. I didn't.
Dispatchers work in a world with its own cadence and its own priorities. A call-taker has a sequence of questions to get through: location, callback number, name. A cop in the field wants the tactical details first: weapon, direction of travel, what's happening right now. Sometimes a caller would blurt out the critical detail and the call-taker would steer them back to the script. Neither side was wrong. They just had different jobs, and each side was frustrated with the other.
I learned a lot in that room. The phones on the emergency and non-emergency lines rang like ordinary household phones, all day, and the stress in that room was real. I pushed for an automatic call distribution system, kept a proposal in my back pocket for the chief, and kept telling anyone who'd listen that dispatchers are really the first first responders. Eventually our IT department found a grant, we got it installed, and the room got quiet enough to think. It's a small thing, but I learned that the right tool for dispatchers pays off everywhere downstream.
The jail had its own turnaround story. In early 2016, before I made captain, the work in both the jail and dispatch earned me the Scottish Rite Officer of the Year award for CVPD. I always felt a little awkward about that award, because the people on the line are the ones who really deserve recognition. But I'll say it plainly: those years were the hardest part of my career, and the part I'm most proud of. They are also where the idea for Live911 was born.
As the new manager, I wanted to hear how my people were doing. Our 911 call logger came with a little monitor app that let a supervisor listen in on a call-taker. So I listened. A lot.
What struck me was how much information came over a 911 call. The caller's tone of voice. Background noise. Urgency, or the lack of it. Little details that never made it into the computer-aided dispatch system, because no human being can type that fast. Then I'd hear the same incident go out over the radio, a minute or two later, boiled down to a few lines. Often there was a big gap between the picture you got from the 911 call and the picture you got from the radio.
It's nobody's fault. It's the system. One person hears the entire emergency, types a summary into CAD, and a second person who never heard the caller reads that summary and puts out a short version on the radio. It's a game of telephone, with time delay and data decay at every step. That's a phrase I coined, and I still use it. I started calling it the 911 gap, and I kept thinking: what if the officers could just hear the call themselves, in real time, in the car?
There was one incident that turned that hunch into a conviction for me. On the morning of February 26, 2014, Martha Huffman was taking out the trash at her home on Twin Oaks Circle in Chula Vista when her ex-husband, Rand Huffman, attacked her with a brick. Neighbors heard her screams and called 911. Afterward I sat down and listened to the 911 call and the radio traffic together. A neighbor calmly described a man beating a woman, and the call-taker was calm and professional. But meanwhile the radio was tied up with lower-priority traffic, and it took time before the call hit the air. An officer who happened to be nearby overheard the assigned units and got there, and Rand Huffman was arrested at the scene. The department celebrated the arrest, and they should have. But I was stunned at how long the information took to travel.
Martha Huffman never recovered. She died of her injuries on April 13, 2014. Rand Huffman later pleaded guilty to first-degree murder, and in July 2015 he was sentenced to 25 years to life in prison. Her case has stayed with me ever since.
So I built a recording that played the 911 call and the radio traffic together, in their real time relation to each other. Then I got in my car about two miles away, played the 911 audio, and drove to the scene obeying all traffic laws, no lights, no siren. By my test, I arrived at about the moment the call was being put out over the radio. An officer hearing that call live could have been there minutes sooner.
I played that video for leadership. They were mildly interested. I got a reaction I would hear many times over the years: if it were such a good idea, someone would have built it already. And then nobody mentioned it again.
I didn't let it go. In May 2015 I started California POST Command College, a futures course for law enforcement leaders that teaches you to study emerging trends and plan for where policing is headed. I was in Class 59, which met for about 18 months in one-week sessions in the San Diego and Del Mar area and graduated on September 14, 2016, a few months before I made captain. The main objective of the course was to research a specific technology or trend, write an article on how it would change the future of policing, and get that article published. I chose streaming 911 audio to officers in the field.
I ran a focus group on the idea and worked through the benefits, the risks, and the downsides. I imagined a system that would push 911 calls to officers based on the caller's location, so an officer could say, in effect, give me every call within a mile of me. In the article I called the concept RAPID, short for Rapid Assistance for People in Distress. The article opened with a story about a rookie officer and his training officer listening to live 911 calls on an app in their patrol car. They hear a robbery victim describe a lifted truck covered in stickers heading north on Third Avenue, and they spot the truck before the robbery has even gone out over the radio. The caller got the color wrong, but the stickers gave it away. Those are the kinds of details that never make it into CAD in time.
The piece was published in Police Chief magazine in January 2017, by which time I was a captain, under the title "Closing the 9-1-1 Gap: How Streaming 9-1-1 Audio Can Transform Emergency Police Response." Police Chief is the most prestigious publication in our profession, and I was proud to get my article published there.
Then I went looking for someone to build it. I reached out to a dispatch analytics company and to several of the call logger companies. They already had monitor apps. It felt like they were one software tweak away. Nobody was interested. So the idea went back on the shelf, and it would sit there until drones gave it a reason to come back.
While I was carrying the 911 idea around, Chula Vista was starting a drone program, and I wasn't the one who started it. That credit belongs to Vern Sallee, with Lieutenant Dan Peak. Around 2015, the department received some free drones from SDG&E, the local power company, and Vern decided to build a team the right way.
Vern took a crawl, walk, run approach. He pulled in stakeholders from around the department and brought in Deputy City Attorney Carol Trujillo. The team studied the ACLU's white paper on drones and privacy and built those concerns into the policy. Then they posted the draft policy online, set up an email address, and invited the public, including the ACLU, to comment. Surprisingly, the comments were mostly positive. People understood the value of drones if they were used properly. Vern also went to the community and to city council leaders for their support.
I can't overstate how important that groundwork was. Without community trust, nobody was going to accept something as new as DFR a few years later. Vern's approach became something of a gold standard for standing up a responsible drone program. It later became the basis for the Five Cs, a roadmap for public safety agencies written mostly by Brendan Groves, working with Charles Werner of DRONERESPONDERS.
By around 2016 the team was flying the traditional way: search and rescue, crash reconstruction, evidence. An officer would get to a scene, decide a drone would help, and call for one. Useful, but reactive and not very time sensitive.
At the very end of 2016, I promoted to captain. Roxana Kennedy had just become chief, and I was her first promotion. With the new assignments, we shuffled collateral duties. Everyone in the building knew I was into drones, so I asked Vern for the drone team. He said sure. I became the UAS Team Commander, and my goal was to make the team truly operational, so that sergeants and watch commanders would think to ask for a drone when they needed one. I also tell on myself here: as a new captain, one of the first things I did with my captain's credit card was buy a DJI Spark, hoping it would be the purpose-built indoor drone I'd been waiting for. I flew it around the command staff room and the chief's office. That pretty much sealed my reputation as the drone guy.
In September 2016, in neighboring El Cajon, officers responded to a call about a man in a mental health crisis. His sister had called for help. The call had been holding for about 50 minutes. When officers arrived, the man pointed something at them that they thought was a gun. It turned out to be a vape device. Alfred Olango was shot and killed, and the community reaction that followed was intense. Roxana Kennedy went out with our mobile field force to help El Cajon during the unrest. Around the same time, officers were being ambushed in other parts of the country, including in Texas.
When she came back, the message to us was simple: put your heads together and find a better way to respond to calls. It seemed wrong that a 50-minute-old, non-criminal call could end that way.
From my dispatch days, I knew how much work could be done before officers ever arrived. As a lieutenant I used to read the holding calls in CAD and phone the callers myself to see if I could resolve them without sending anyone. And I kept thinking about scouting. The military has always sent someone up the hill to see what's on the other side. In policing, we were sending two officers call to call, in the blind.
I talked the idea through with Lieutenant Kenny Hines, who thought it was good, and then teamed up with Lieutenant Eric Thunberg, who iterated on it and really took hold of it. It was a team effort. With approval from Chief David Bejarano and Captain Roxana Kennedy, who was second in command at the time and would become chief at the end of that year, we ran a formal test of about three months in the fall of 2016. I named it the SARA Project, after the community policing problem-solving model: Scan, Analyze, Respond, Assess.
The idea was to put one plainclothes officer in an unmarked car as a dedicated resource. That officer had two jobs. First, work the holding calls in CAD and clear what could be cleared. Second, on hot calls, get there before the uniforms, look the situation over without being seen, and tell responding officers what they were walking into. We gave the unit its own call sign, SARA 1, so that when officers heard it on the radio they'd know to hold back and let the scout work.
It worked, in the sense that it proved the concept. On at least one call, the SARA officer got there first and saw something involving a knife that otherwise would have been missed. But it was effective and inefficient. One officer in one car can't beat the ground units across a whole city, and that officer isn't driving much faster than anyone else. The project faded out. It was a relative success, and it planted the question that became DFR: what if the scout didn't have to drive?
In early 2017, I was online looking for indoor drone solutions and came across a small company called Cape, short for Cape Aerial Telepresence. I sent them an email asking about indoor drones. Kabe Termes, who was running operations at Cape, wrote back and said he'd love to come demo the product.
Kabe and his colleague Norm Serrano flew down from the Bay Area and demoed it upstairs at the department for me and Lieutenant Chris Kelly, who was helping me with the drone program. Cape had built software that let you fly a drone remotely over the internet from any connected device, with geofencing to keep it inside a safe box in the sky and the ability to stream the video to other people. They pitched it as a streaming solution for public safety, but honestly, nobody, including Cape, had figured out how public safety would actually use it.
For traditional drone work, it didn't make sense. You'd need someone on scene with the drone and someone back at the station flying it. That's just extra people.
Then it clicked. SARA had proven the value of getting trusted eyes on a scene first. Cape made it possible to fly a drone from a rooftop, remotely, to that scene. Put those together and you have a drone launching ahead of the officers, flown by an experienced cop sitting safely in a room, who can see what's happening and talk the officers in. That was the moment DFR was born. It resonated with Kabe right away.
I want to be fair about this. I don't think it was a completely unique idea. A lot of people had imagined a drone in a box that would launch and tell you which side of the freeway the crash was on. What hadn't existed was a practical way to do it every day in patrol. The mix of our SARA use case and Cape's software made it practical for the first time.
Kabe, Norm, Chris Kelly, and I kept talking and emailing about it. We thought we could pitch Chief Kennedy, who had a reputation as a trailblazer, on a small beta test. Then, in October 2017, the White House, the Department of Transportation, and the FAA announced the drone Integration Pilot Program, the IPP. It invited public agencies to partner with private companies on new ways to use drones, so the FAA could learn how to safely bring them into the national airspace. I like to joke that the White House must have been listening in on our calls. It was exactly what we were already trying to do.
We wrote up a concept of operations. Kabe and Cape's CEO at the time, a young guy, the typical startup type, came down and pitched the idea to Chief Kennedy, the city manager, and others from the city, including Eric Crockett, Chula Vista's director of economic development. They believed in it, and it fit the city's smart city strategy.
At first we planned to apply as Chula Vista alone. But the City of San Diego was also applying, with big players like Qualcomm, Intel, AT&T, Uber Eats, Matternet, and AirWorks. Eric Crockett had a connection in San Diego and made the case that we should join their team instead of competing with it. Chief Kennedy wasn't thrilled at first. She wanted it to be ours. I always smile at how my idea became our idea, but that's how it should be. It started as Kabe's and my idea, and it was great to see other people own it. We sent our draft concept of operations to San Diego's team, they folded it into their submission as the public safety use case, and it went in.
With the bid submitted, I did the retirement math and decided it was time. I retired in May 2018 after 27 years.
About two weeks later, the announcement came: San Diego's team had been selected as one of the ten IPP sites nationwide, with Chula Vista's drone concept as the public safety piece. Timing is funny like that.
After we were selected, Kabe had some of his people, including Omar Mercado, who still works with me today at Flock, go down to Ensenada in Baja California and work with the police chief there. They set up a form of the concept, flying an Inspire with the police department. The chief even signed a document granting them airspace authorization, which we found pretty funny, since airspace is a federal matter on our side of the border. They caught a group of burglary suspects after launching off a roof, and you could watch the chase on the video. I got permission to cross the border and went down to evaluate what they were doing. It wasn't DFR the way we would do it. They used the drone more for proactive patrol, scouting ahead of officers in a truck, and the equipment had real limits. But it was the first time I got to see something like it in person.
Vern Sallee took the drone team back. He had started it, I had taken it over, and now it was his again. Since I'd done a lot of the groundwork with Cape, he invited me back as a consultant. Chula Vista is a mid-sized department and everyone there already had a full-time job. So for about a year I worked alongside Vern, Chris Kelly, Don Redmond, Jim Horst, and others to stand the program up. We evaluated launch sites, did public education and gave a lot of tours, wrote policies, helped with the training program, and worked on the waivers. That year is really where I cut my teeth on drone operations. I thought I was retiring. It didn't happen, thankfully.
On August 10, 2018, before our program launched, Kabe and Chris Rittler, who was Cape's CEO by then, were impatient to get the concept moving and set up a demo with San Diego Fire. I invited myself and went out to watch them fly to a house fire, a mock one if I remember right. Fire never really pursued it. In fact, some of the fire folks didn't want a drone called a first responder at all, since it wasn't a person.
Not long after that, Kabe, Chris, and I sat down for lunch in San Diego to talk about how to get Chula Vista moving and what to call this thing. We tossed around a lot of names. I suggested Drone as First Responder. At first the reaction was, well, that's just what it is, it's common sense. Exactly, I said. And honestly, the fact that the fire guys didn't like the term made me more determined to use it. Cops didn't care about that. It made sense. They agreed.
On September 20, 2018, I wrote an email to Vern Sallee and the rest of the team. I explained that in the age of Google, this use case needed one consistent name, so that anyone searching for it would always find their way back to Chula Vista. I asked that whenever possible we call it Drone as First Responder, DFR for short. Not drone first responder. Not drone as a first responder. Drone as First Responder.
I wasn't sure at first whether the name would catch on. But people needed some name for what we were doing, and over time that one stuck. Chief Kennedy and the UAS team embraced it. It got so official that they painted it on the wall of the DFR room and on the rooftop launch platform. The first time, the painters wrote "drone as a first responder." Chief Kennedy, perfectionist that she was, had them fix it.
On October 25, 2018, Chula Vista launched its first official DFR mission from the roof of police headquarters. It was the first official DFR program.
Here's how it worked in those early days. A Part 107 licensed pilot stood on the roof. That pilot was officially in command. They kept the drone ready, swapped batteries, watched the airspace, and could take over the controls instantly if a manned aircraft came near. Downstairs, in a room set up like a small real-time crime center, sat a sworn officer we called the teleoperator. Using the Cape software, the teleoperator could take off with a keystroke or drop a pin on a map and let the drone fly itself there, inside the geofence.
The teleoperator didn't need to be an expert pilot. The skill that mattered was police work. They watched the holding calls in CAD and the radio, decided where a drone would add value, and launched. Once overhead, they got on the radio and told officers what they were seeing: hurry up, slow down, come in from the other side. I've described it as being like a conductor leading an orchestra. That person isn't in harm's way, isn't running red lights, and can think clearly.
For the first several months we had to keep the drone within visual line of sight, which meant about a one-mile radius around headquarters. Even so, there were enough calls in that area to prove the value. One call from those early days stands out. The teleoperator flew to a call that was holding, one we didn't even know was urgent, and saw a woman in a car trying to run over a man on a motorcycle in a domestic violence situation. Cops were sent right away. That was one of our first real aha moments.
DFR also turned out to be a resource management tool. A lot of calls, a fight in a park, kids tagging, someone loitering, normally require two officers to drive across town just to find nobody there. With the drone overhead, the teleoperator could see the scene had cleared, call the reporting party, and clear the call. That frees those two officers for something urgent.
To scale, we needed to fly farther. A one-mile radius covers about three square miles. A three-mile radius covers roughly ten times that. Ten times the area means ten times the calls you can help with, and that's what makes it worth paying for a pilot on the roof.
After about seven months of flying safely inside a mile, I worked with Kabe Termes and Ben Kroll of Skyfire Consulting to write the safety case for a beyond visual line of sight waiver. The Chula Vista team provided the flight data that backed it up, and the IPP gave us direct access to help from the FAA.
In March 2019, the FAA approved it. Chula Vista could now fly out to about three nautical miles from the launch site. The rooftop pilot no longer had to see the drone itself, only the airspace around it. Chula Vista became the first agency in the country approved to fly beyond visual line of sight to respond to calls for service in an urban environment, and the first to do it as Drone as First Responder.
By the fall of 2019, there was a second launch site, on a medical facility about a mile south of headquarters, and those two sites covered about a third of the city's area, but that third produced roughly 70 percent of the priority calls. The team was flying a single 40-hour shift, Monday through Thursday, and getting eight to fifteen flights a shift on DJI Matrice 210s with a 30x zoom camera. That zoom mattered. From 250 feet or more, the teleoperator could read a license plate or see what was in someone's hand without dropping low and making noise.
We were transparent about all of it. Chula Vista posted daily flight statistics online, and every flight's path, time, and reason was posted for the public to see. The policy was to fly only to actual calls, never just to look around. In community meetings, we got essentially no pushback. Some residents actually wanted us to use the drones more.
I used to take five-mile walks around Alpine listening to podcasts about public safety drones. In 2019, on one of those walks, I heard Chief Charles Werner interviewed on the Drone Radio Show. He was talking about Chula Vista's program and DFR, and he encouraged people to reach out to him. So I did, and introduced myself as the guy who'd come up with the idea.
Charles is a retired fire chief who founded DRONERESPONDERS, which grew into the largest nonprofit advancing public safety drone operations. We hit it off and became good friends, and he became one of the leading voices for DFR across the country. Over the years we've teamed up on a lot, including the waiver work I'll get to in a few chapters.
Once DFR was flying, my old 911 idea came off the shelf with a new purpose. If a DFR teleoperator could hear the 911 call as it was happening, they could launch at the first sign of trouble, before the call was even entered into CAD. The drone could be overhead before the old system had even started its clock. Today people call that negative response time.
At some point I remade my old drive-test video using a drone. I went up on the CVPD roof with my own Mavic, played the same 911 audio, and launched at the start of the call. The scene was close to the station, and I was overhead in about a minute. That, to me, was the whole picture: see and hear, as early as possible.
After I retired, I was telling people in the drone world about the idea, and someone told me they knew exactly the company I needed to call. That was HigherGround, which makes call logging and recording systems for 911 centers. I reached out, and the CEO, Terry Ryan, sent Tom Goodwin, a vice president there, to meet with me. Tom and I met twice in Coronado. He believed in the idea and redirected his engineers to build a minimum viable product. Terry gave him the room to do it.
Now I needed a place to test it. I went to Don Redmond, who was then the lieutenant managing our 911 center, and to Eric Wood, our IT director. Don had been through Command College too and understood exactly what I was after. Eric understood the technology. I'll be honest: I didn't take it to the whole command staff right away. I was worried it would get shut down as one more good idea at a time when everyone was supposed to be focused on DFR. So we started quietly, in the 911 center and the DFR room, and the DFR pilots became the first beta users. That also let us test it without putting it in patrol cars yet, because people were nervous about how officers would react to hearing raw 911 calls.
In July 2019, the officers assigned to DFR began beta-testing the software. Within about a month it had its first success story. A call came in about a postal employee threatening to come to work with an AK-47. The drone operator heard it live and launched before the call was even in CAD, found the subject within a couple of minutes, and guided officers in. Fortunately, the man was unarmed.
DFR pilot Matt Hardesty became one of its biggest believers. He once heard a 911 call where the caller hung up, and when dispatch called back, a woman said it was an accident. But he'd heard someone yelling in the background, and it didn't sound right. He flew over and saw a young girl come out of the house with a large knife and start scratching a car in the driveway, followed by a confrontation with her mother. At that point there was no call for service and no officers on the way. He put it out, units went Code 3, and nobody got hurt. In those early days the system was down a lot for updates, and Matt didn't even want to fly DFR without it, because waiting for CAD felt too late.
In another case, at a local taco shop, a caller reported a man waving a gun around. Officers heard the call live and a drone launched. Matt saw the man pick up the gun-shaped object and use it to light his cigarette. It was a lighter. He told the responding officers right away, and they adjusted their approach. Those are the moments that keep me doing this.
As for the name, I'd called it RAPID in my article. But Don Redmond and his team just called it Live911 when they talked about it. When Tom Goodwin was looking for a product name, we asked Don, and he said, we just call it Live911. It's a great name because, like Drone as First Responder, it simply says what it is.
The dispatchers had worries at first. They feared officers would second-guess them or start coaching them on the air. The opposite happened. Officers heard how hard call-taking really is and gained a new respect for dispatchers. Radio traffic actually went down, because officers could hear the call-taker already asking the questions.
In mid-May 2020, Chula Vista put Live911 in every patrol car, the first agency in the country to livestream 911 calls to officers in the field. Each officer could set a radius, say half a mile, so they only heard nearby calls, and the integration with RapidSOS showed them on a map exactly where the caller was standing. HigherGround launched Live911 nationally on July 15, 2020, with Clovis, California, and Polk County, Florida, finishing beta tests. Clovis had visited our DFR room, seen Live911 in action, and wanted it for its own officers. HigherGround later received patents on the technology that name me as one of the inventors.
Chief Kennedy later wrote that, for the first time in ten years, Chula Vista met its response goal for priority one emergency calls, and she credited Live911 as instrumental. Live911 was also part of three separate life-saving awards at Chula Vista, and those came before any DFR-related life-saving awards. If DFR is the eyes, Live911 is the ears.
During my consulting year, I got to know Skydio. When I first saw a Skydio drone fly itself and avoid obstacles on its own, I realized those people had the DNA for the future of drone operations. Automation was exactly what DFR needed to scale. I started working with them as a consultant, and in November 2019 I joined Skydio full time as Head of Public Safety Integration.
My job was to build a bridge between Skydio and public safety, talking with agencies every day about their pain points and bringing that back to the engineers. Kabe Termes ended up joining Skydio too, and for a while I was actually working for him, which felt like a small full circle. Cape itself was later acquired by Motorola. I stayed close to Chula Vista the whole time.
At Skydio I also tried to put words around what makes DFR work. I broke it down into five problem sets I call the 5 As: the aircraft, aircraft management (launching, landing, and changing batteries), airspace awareness, autonomy, and approvals, from both the FAA and the community. You have to solve all five. I spent about four years at Skydio, and that framework has followed me ever since.
At Skydio, I kept struggling with what "beyond visual line of sight" really meant. When the FAA wrote the rule, were they thinking about a drone flying so far away you couldn't see it? Or did they also mean a drone flying behind a house or under a tree canopy, low to the ground, where it's no danger to manned aircraft? Ask three FAA folks and you'd get three answers. Meanwhile, public safety was doing that kind of flying every day: looking in windows on a barricaded suspect, checking a river bottom for a missing person. Nobody was sure if it was legal.
My idea was simple. What if we promised the FAA we'd stay low, down around the height of the trees and buildings, and used a drone that could see and avoid obstacles on its own? Then you could fly out of sight safely and get it done in the first moments of an emergency.
I called Charles Werner to ask if anyone had tried this. He pointed me to Chris Sadler with York County, Virginia, who had already gotten a waiver along those lines. Chris was technically the first. Chris generously shared his safety case, but it required multiple drones and ADS-B and was too complex to set up when a suspect is running around a corner. His safety case used the phrase low altitude, close proximity. I flipped it around to Close Proximity Low Altitude, or CPLA, because I wanted a single term for the operation, just like DFR.
When I mentioned it to Kabe, he was focused on getting approval for one pilot to fly multiple drones and didn't want to distract the FAA with a competing concept. Fair enough. Then Brendan Groves joined Skydio to lead regulatory and policy work, after leading the drone program at the Department of Justice. He liked the idea and ran with it. Together with the San Diego IPP team and Chula Vista, we put together a safety case that relied on Skydio's obstacle avoidance and submitted it through the IPP.
We spent months on calls with Mike O'Shea and Tony LaPorte and others at the FAA, with Charles Werner helping make the case for how important and how safe this was. In June 2020, the FAA granted Chula Vista what it called a tactical beyond visual line of sight waiver. The FAA wrote its own version and did not require obstacle avoidance, so any drone could be used. I'll admit that was less safe than what I proposed, but I understood they didn't want to pick winners.
Meanwhile, Brandon Karr in Pearland, Texas, who I work with today, had been building templates to help agencies get a Sadler-style waiver. Pearland got its tactical waiver right after Chula Vista, by my recollection the very next day. Charles and DRONERESPONDERS put out templates explaining how to get it, and within a few months more than a hundred agencies had it.
That waiver held up for years. In the spring of 2025, I told a panel that we had just had a breakthrough with the FAA, maybe the point where we'd talk about DFR before and DFR after. In 2025 the FAA's Flight Standards team rolled out a templated waiver for beyond visual line of sight flight up to 200 feet. An agency simply emails Flight Standards to request it. There's no CAPS application and no COA, and approvals come almost immediately. Today well over a thousand agencies have it, maybe two thousand. The tactical waiver has largely been folded into that one. It was one more step along the way.
I remember when a thousand missions felt like a milestone. Here's roughly how Chula Vista's DFR program grew, from what was being reported along the way:
By October 2019, about a year in, the program was at roughly a thousand missions. By then the drones had assisted in more than 120 arrests and cleared more than 200 calls without sending an officer, with an average response time of about two minutes.
By May 2020, it was about 2,500 missions, and the team was flying seven days a week.
By late July 2020, it was crossing 3,000, with 12 to 18 missions on a typical day.
By the fall of 2020, it was past 3,500 and heading toward 3,800.
By January 2021, it was over 4,000. Around 20 percent of the time, the drone got overhead and found that no ground units needed to respond at all.
In March 2021, according to Chief Kennedy's later testimony to Congress, Chula Vista became the first agency authorized to launch from anywhere in the city, and it added launch sites to cover the whole city.
By 2024, Chula Vista announced its 20,000th DFR mission. Today the program has flown well over 30,000 missions.
Other agencies followed. Rahul Sidhu started the second DFR program, at Redondo Beach. Tijuana's police department visited and set up its own version. Clovis's police chief, I believe, wrote a Command College paper on the concept. Brookhaven, Georgia, had its city council fund a program. Over time DFR spread to agencies across the United States and to Mexico, Canada, the UK, Europe, and Australia.
Through all of this, most DFR programs still ran the way we did on day one, with people on the roof. I call that DFR 1.0. Everyone knew that to really scale, we needed DFR 2.0: docks that launch, land, and swap batteries on their own, and detect-and-avoid systems that let the FAA approve flying with no one on the roof. Pearland, under Brandon Karr, ran the first version of DFR 2.0, though with real limits.
Rahul Sidhu is a cop, a drone pilot, a technologist, and a founder. After starting the Redondo Beach program, he founded Aerodome. After seeing Aerodome demonstrated at a DFR 2.0 conference in Chula Vista, I left Skydio in March 2024 and joined Rahul and his team. Aerodome was a small startup, but it was doing DFR better than anybody. I started as vice president of strategic initiatives and then became vice president of professional services. My job was helping agencies get their FAA waivers, write policies and protocols, and explain DFR to their communities: what it is, why it helps, and what the agency will and won't do with it.
In October 2024, Flock Safety announced it was acquiring Aerodome. I came over with the acquisition and became Flock's Director of DFR Consulting.
My job really hasn't changed much. I still help agencies with FAA waivers, policy, training, and community engagement.
Chapter 20: Where DFR and Live911 Are Going
When I started, nobody was building docks, because flying beyond visual line of sight wasn't really allowed. Now those pieces are finally coming together: docks, detect-and-avoid systems, simpler FAA waivers, and a proposed new FAA rule, Part 108, that should make it more turnkey still. Several companies, including some of my friends and former colleagues, are competing hard in this space. Kabe Termes is a vice president at Skydio, leading its solutions engineering team. Vern Sallee retired from Chula Vista in 2020 and went to Axon, where he works on its drone strategy. As someone at Skydio put it, the DFR wars have begun. Competition is how the best solutions get to agencies, and the real winners will be the communities.
Live911's idea has gone the same way. What once felt like my odd little idea is now an industry standard. Several products now stream 911 audio to responders, and when they describe what they do, they say you're listening to live 911 calls. I take that as a compliment.
The next frontier is bringing dispatch and the field closer together in both directions. Streaming 911 gave officers a window into dispatch. Now drone video is giving dispatchers a window into the field, so they can see the scene they've been talking a caller through. Real-time crime centers and DFR are merging into one new role that sits between the 911 center and patrol, which is exactly the position I imagined for the SARA officer and the teleoperator years ago. And the same ideas are moving into private security, with drones protecting campuses, parks, and business properties.
After all these years, my core belief hasn't changed. DFR is not about cops getting drones quickly to the scene. It's a way for drones to get cops quickly to the scene, with an experienced decision-maker virtually there first, seeing the big picture and out of harm's way. The goal has always been to know what's behind the next door before anyone has to open it. People credit me with the line that drones are a tool to see the future, and I still believe it.
None of this happens without people willing to grab a torch and run with it: Chief Roxana Kennedy, Vern Sallee, Dan Peak, Eric Thunberg, Kabe Termes, Chris Kelly, Don Redmond, Eric Wood, Matt Hardesty, Ben Kroll, Jim Horst, Charles Werner, Brendan Groves, Chris Sadler, Brandon Karr, Tom Goodwin, Terry Ryan, Rahul Sidhu, and many more. I'm grateful I got to run alongside them. And I'm not done yet.
We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.