This is the 5th installment of my interviews of early-stage companies impacting the emerging technologies ecosystem.
I had the pleasure of interviewing three Georgetown University classmates, John Andrzejewski, Lucas Raskin, and Eli Kerstein who founded Guardian RF last year after winning a defense hackathon. This was especially meaningful as I serve in Georgetown’s Adjunct Faculty in the Graduate Cybersecurity Risk Management program.
The students already have a record of success as they have deployed their systems on the frontlines in Ukraine, helping protect warfighters from explosive FPV and deadly drones.
Their innovative technology can detect drones from up to 5 km away, accurately pinpoint their location on a map, and identify the operator’s position. Unlike many drone detection systems, Guardian RF’s sensors can detect modified and non-remote ID-compliant drones, ensuring comprehensive coverage against evolving threats.
Much of their current work focuses on securing homeland security critical infrastructure, including energy assets and data centers running sensitive government workloads. Drones present a growing threat by bypassing traditional security measures to deliver payloads, conduct surveillance, or disrupt operations without leaving a trace. These airborne threats are becoming a key concern in both the physical and cybersecurity domains, especially as they target critical infrastructure.
Chuck Brooks (CB): What is Guardian RF?
GRF: Guardian RF is a drone detection company founded by three physicists with a background in signal processing research. During a Summer Batch at Y Combinator, they deployed their drone detection platform with Ukraine’s Territorial Defense Brigades in operations at the zero line.
The company is built around the principle that countering drones must be as cost-competitive as drones themselves. Guardian RF’s detection technology helps security teams address threats from UAVs—whether it’s contraband delivery, unauthorized surveillance, or disruptions of crowded events—safeguarding public safety and vital industries.
CB: Can you share the story behind Guardian RF? What inspired you to enter the drone detection market?
GRF: The three of us met in the Physics Department at Georgetown University, where we had worked on signal processing projects together. Our entry into the drone detection space happened somewhat by chance—we were encouraged by some peers to participate in a defense tech hackathon in El Segundo, CA. That event put the drone threat on our radar, and the solution we developed—a handheld drone detector—was ultimately selected by the Ukrainian Ministry of Defense as the most promising. This early validation led to our first angel investment, followed by a pre-seed round from Y Combinator.
From there, we split our team between San Francisco, where we participated in YC’s accelerator program, and Ukraine, where we deployed early versions of our system with territorial defense brigades. Since then, we’ve raised a seed round, and started deploying our systems here in the United States.
CB: What brings your focus to Homeland Security?
GRF: In Ukraine, we saw firsthand what soldiers called “warehouses of shame”—stockpiles of expensive, over-engineered American-made sensors and ISR platforms sitting unused. Many of these systems were too bulky or power-hungry for frontline use, so they were being cannibalized for parts. Soldiers would strip them for software-defined radios and compute, then repurpose those components to build smaller, battle-borne detectors. This revealed a fundamental truth: the drone threat is entirely defined by economies of scale. Any effective counter-drone system—detection, tracking, and ultimately mitigation—needs to be cost-competitive with the drones themselves.
The problem isn’t just technical; it’s systemic. Most drone defense solutions have been developed under the assumption that this is solely a military issue to be tackled within traditional defense procurement processes. But the threat is broader and highly asymmetric, affecting everything from our nation’s densest urban centers to our national parks. Any untrained actor, with as little as a few hundred dollars can deploy the cutting edge of weaponry to disrupt American public safety. Counter-drone technology must be scalable, affordable, and adaptable to real-world constraints.
CB: What differentiates your system from other solutions on the market? Can you explain how your drone detection technology works?
GRF: We take a service-based approach to drone detection, focusing on what users actually need rather than upselling individual units. A six-figure drone detection system might be an easy sell to the Pentagon, but businesses, police departments, and critical infrastructure operators can’t afford systems designed for government-scale spending. Law enforcement and security need cost-effective, reliable solutions to detect and mitigate these threats—without the unnecessary complexity of radar or high-end camera systems.
Our system works by passively detecting the RF signals drones use to communicate. Every drone has to stay connected to its operator, whether for control or for video transmission, and that connection leaves a signature in the electromagnetic spectrum. Instead of trying to actively ping or track drones like radar does, we simply listen for those signals.
We do this in a few ways. One is Remote ID monitoring, which works like a digital license plate mandated by the FAA for all drones models commercially available in the United States. We intercept and decode these identifying signals, though Remote ID isn’t always enabled, so we also analyze the video link itself. Drones use frequency hopping spread spectrum (FHSS), meaning they rapidly switch frequencies in a predictable way. This hopping pattern is unique to each UAV and acts like a fingerprint. By tracking it, we can identify and classify drones in real-time. We also look at the shape of the spectrum—how a drone’s signal visually appears on a spectrogram. Different drones have distinct RF characteristics, and by recognizing these patterns, we can not only detect drones but also determine their type and origin.
To localize drones and their operators through time delay of arrival (TDoA), we use a mesh network of synchronized sensors that work together to pinpoint the source of a signal. The challenge with TDoA is that it requires precise time synchronization across multiple sensors. Since radio waves travel at the speed of light, even a nanosecond difference in signal arrival times can mean meters of error in location tracking. To solve this, our sensors synchronize their internal clocks using high-precision timing protocols, ensuring they are in lockstep with each other. Once synchronized, each sensor listens for the drone’s signal and records the exact moment it arrives. By comparing these timestamps across the network, we can triangulate the precise location of both the drone and its operator. This method allows us to track threats in real-time, even if the drone isn’t broadcasting a Remote ID.
CB: Could you highlight some recent trends or emerging threats in the drone landscape that public safety officials should be aware of?
GRF: One of the biggest developments happened just last month—on January 13, 2025, DJI announced that it will no longer enforce geofencing on its drones. Previously, DJI drones had built-in restrictions that prevented them from flying in sensitive areas like airports or stadiums. With this change, the responsibility now falls entirely on the operator, meaning anyone—whether well-intentioned or malicious—can easily bypass restrictions that once required hacking or modification. This will make unauthorized drone incursions significantly more common, and public safety teams need to be prepared to detect and respond in real-time.
Another growing concern is the use of drones to target energy infrastructure. We’ve already seen examples of lone-wolf actors firing at substations with rifles, but drones remove much of the personal risk as they offer anonymity. An individual can now conduct surveillance, drop small payloads, or even interfere with operations from a safe distance—making this kind of attack harder to predict and prevent. This threat is anything but theoretical. Lone-wolf quadcopter bombings at substations in the American Southwest last winter proved it’s real, and it’s only a matter of time before we see more sophisticated attacks.
Then there’s the psychological impact of drone threats, particularly in public spaces like amusement parks or stadiums. A lot of people assume that the risk of a drone attack means something extreme—like someone dropping anthrax over a stadium. But the reality is that it doesn’t take a chemical weapon or explosive to create chaos. If a drone were to disperse something as harmless as baby powder or flour over a packed stadium, it could trigger mass panic, stampedes, and widespread disruption.
CB: What are the relevant legal and regulatory challenges that organizations face when implementing drone detection? How do you see these regulations evolving over the next few years?
GRF: Our system is designed to be fully compliant with existing laws while giving security teams real-time intelligence on where a drone is and, more importantly, where the operator is located. Tracking down the drone operator in real-time is so important because it is the most unilaterally legal and straightforward way to interdict a drone threat. Just as critical is knowing whether the same operator has previously entered restricted airspace or is properly registered, which our platform automatically tracks through a built-in offender registry.
Jamming drone signals, for example, is illegal under FCC regulations since it interferes with the broader radio spectrum and could disrupt unintended signals. Local law enforcement’s authority to shoot down drones is legally complicated as the FAA controls U.S. airspace and federal law classifies drones as aircraft. While a few federal agencies have counter-drone authority, local agencies lack a clear legal framework, making operator interdiction the most straightforward approach.
What’s happening now is that many security teams have been relying on outdated or ineffective solutions—one of the biggest being DJI Aeroscope, which was widely used for drone tracking but only worked on DJI drones and, even then, fails to detect modified or masked signals. Now that DJI has discontinued Aeroscope, many organizations are stuck holding a defective product.
First, when it comes to regulations that require drone security, we’re seeing a growing number of industries mandated to take action. The NFL now requires all stadiums to have some form of drone detection at the league level, recognizing that drones pose both security and game-day disruption risks.
Regulatory frameworks to watch include NERC CIP (North American Electric Reliability Corporation Critical Infrastructure Protection) and CFATS (Chemical Facility Anti-Terrorism Standards). While NERC CIP has traditionally focused on cybersecurity standards for the energy sector, there’s a growing push to address physical threats—such as drones—within its scope. CFATS, which governed security at chemical plants and high-risk facilities, expired in 2023 and remains in limbo. However, there is widespread industry recognition that these facilities remain prime targets for drone attacks, and some facilities are proactively implementing drone detection ahead of potential new mandates.
Drone security today is where cybersecurity for critical infrastructure was a few years ago—the risks are clear, but the regulatory environment is still catching up. Right now, we’re likely two years into what will be a decade-long process of regulatory development, as policymakers work CB: to keep pace with the evolving threat landscape.
CB: What kind of training or expertise is needed to operate your systems effectively?
GRF: Our system is designed to be as simple as possible to deploy and use, requiring no specialized training. Installation is straightforward—just connect the device to power, and it starts working immediately. From there, users can log into our secure server with the credentials we provide, and the drone detection process is fully automated. Everything is accessible through a web-based interface that works on any device, similar to logging into a website on your phone. For organizations monitoring multiple locations, we provide an enterprise dashboard that consolidates all drone activity across various sites, allowing users to set geofences with a single click, define alert thresholds, and receive notifications directly to their phones if an unauthorized drone enters a restricted area.
Some organizations prefer not to use our GUI and instead stream the data into their existing Security Information and Event Management (SIEM) systems, allowing them to monitor drone activity alongside other security alerts. The system is designed to be flexible, whether a user wants a completely hands-off experience where alerts are automated or a more customized setup where they can whitelist drones, fine-tune detection zones, and adjust geofencing settings as needed. It is built to work out of the box with no hassle, while also providing advanced integration options for teams that want deeper control.
CB: What is the process for responding to an alert? What actions should security teams take once a drone is detected?
GRF: Once an alert is triggered, the first priority is verifying the threat and determining the level of risk. Our system provides real-time intelligence, allowing security teams to immediately assess whether the drone is broadcasting Remote ID, what type of communication link it’s using, and whether it has been previously detected. If it’s an authorized or whitelisted drone, the system can filter it out automatically. If not, the next step is to determine its location and intent.
Because Guardian RF tracks the drone’s operator, security teams can act at the source rather than just watching the drone itself. Since mitigation options are limited—security teams cannot legally jam signals or shoot down drones under FAA and FCC regulations—the most effective course of action is locating and interdicting the operator before they can cause harm. Once an unauthorized drone is detected, the response typically follows a standardized protocol. Security teams verify the alert by checking the drone’s Remote ID, signal type, and flight pattern to determine if it is a legitimate threat or an authorized operator. They then track the drone in real-time, using its movement, altitude, and signal characteristics to assess whether it is surveilling an area, delivering a payload, or simply passing through.
The next step is locating the operator using Time Difference of Arrival (TDoA) tracking, which triangulates their position so security personnel can quickly respond. Ground teams are then deployed to the operator’s location to assess the situation. Depending on the intent of the operator, the response may range from a simple warning to law enforcement intervention. Many drone incursions are due to carelessness rather than malice, such as hobbyists unaware of airspace restrictions or contractors using drones for inspections without proper authorization. However, for persistent or intentional violations—such as unauthorized surveillance, smuggling, or repeated incursions—security teams may escalate the response accordingly. The goal is always to resolve the situation efficiently while preventing unnecessary disruption or panic.
CB: How accurate is your system in detecting drones, and what factors affect reliability? What is the effective range, and how do you differentiate between authorized and unauthorized drones?
GRF: Our system is highly accurate, with false positives being nearly nonexistent under normal conditions. In our ongoing commercial pilots, we have never recorded an instance of a false positive or false negative. The only known false positives in our testing have come from sophisticated military-grade spoofing, where a malicious actor broadcasts signals mimicking legitimate Remote ID packets. We counter this by cross-referencing Remote ID with the drone’s video link—if no video link is detected, the system discards the signal.
For false negatives, detection success depends on range, weather conditions, and physical obstructions. In our testing, a drone actively broadcasting within 2.4 km at an altitude of 50–600 meters is nearly always detected, with the only exceptions being extreme interference or signal attenuation from obstacles. We continuously refine our RF front-end hardware, antenna gain, and filtering algorithms to improve reliability, and when properly deployed with a 7-foot antenna, our false negative rate is just 0.8% (measured as a fraction of total drone flight time).
The system distinguishes between authorized and unauthorized drones by cross-checking FAA Remote ID and LAANC authorizations against a client’s internal whitelist of approved operators. If a drone is registered in either system and flying within its approved location and time window, it is flagged as authorized. Otherwise, it is marked unauthorized. Security teams receive real-time alerts with exact drone location, operator data (if available), and flight details, ensuring they can respond appropriately.
CB: How do you handle maintenance or updates to the system, especially as drone technology rapidly evolves?
GRF: We operate on a fully managed service model, meaning everything—hardware, installation, networking, and maintenance—is included, so users don’t have to worry about upkeep. Software updates happen automatically, ensuring the system is always up to date with the latest detection capabilities as drone technology evolves. We handle all firmware improvements remotely through secure over-the-air (OTA) updates, allowing us to push security patches, performance enhancements, and new detection features without requiring any on-site intervention.
On the hardware side, continuity is fully covered—if a device ever has an issue, we replace it free of charge, ensuring uninterrupted service. Unlike traditional systems that require expensive maintenance contracts or manual troubleshooting, our approach eliminates the risk of outdated or failing hardware disrupting drone detection. To ensure consistent operation, we’ve built multiple layers of redundancy into both the hardware and software. Each sensor has a battery backup, so it keeps running even if the primary power source fails. We also deploy multiple sensors in overlapping coverage areas, so if one goes offline, the system still functions seamlessly.
At the software level, we have continuous self-monitoring. Every sensor sends a heartbeat signal once per second, confirming that it’s online and functioning correctly. If a sensor ever stops responding, our system automatically detects the issue and flags it before it becomes a problem. Additionally, we conduct routine test broadcasts—our system transmits a known Remote ID packet to verify that all sensors are correctly detecting and processing signals, ensuring that nothing is missed due to configuration errors or environmental interference. To prevent software failures, we use automated watchdog processes—these constantly monitor system performance and will immediately restart any service that stops or becomes unresponsive. Our system also tracks internal metrics like CPU usage and temperature, allowing us to detect early warning signs of potential failures.
For users, this means they never have to worry about maintenance—the system is designed to run reliably on its own without requiring technical expertise or constant oversight. If something does go wrong, our built-in self-healing mechanisms ensure the system can quickly recover, and if a hardware replacement is ever needed, we handle it seamlessly at no additional cost. This approach guarantees that organizations always have fully functional, up-to-date drone detection without the complexity or hassle of managing the technology themselves.
CB How can Guardian RF be contacted?
GRF: Via our website: Guardian RF
