The Department of Homeland Security’s own test lab put six Blue UAS in front of 26 public safety pilots, and three of the five it brought to Manhattan could not be flown safely there. The Blue UAS for First Responders Comparative Assessment Report, published this month by the National Urban Security Technology Laboratory (NUSTL) in New York, covers three rounds of testing between June 2024 and July 2025. The series started in rural Texas, moved to Miami that November, and ended at Hudson Yards in Manhattan. The Ascent AeroSystems Spirit, the Freefly Astro, and the Teal 2 were excluded from scoring at Hudson Yards because they could not hold position without drifting. Only the Parrot ANAFI USA GOV and the Skydio X10D were rated there.
DHS lists four takeaways at the top of the report: GPS cannot be relied on in dense cities, the link between controller and aircraft is inconsistent, deployment takes too long, and cost is a barrier for many departments. The systems as tested cost between $13,964 and $56,195.
DHS also flew drones from covered nations, the report’s term for Chinese-made aircraft, alongside the Blue UAS at Hudson Yards. Those results are not in the report. Public safety agencies can get them by emailing the lab.
Three of Five Blue UAS Could Not Hold Position at Hudson Yards
A footnote in the DHS comparative assessment report records the Spirit, the Astro and the Teal 2 as unsafe to fly at Hudson Yards because none could hold position without drifting. The test team and safety pilots blamed RF interference and GNSS trouble from the buildings. None of the three was scored.
The Astro would not enter position mode at Hudson Yards and drifted constantly. A safety pilot from the test team took control and tried to switch it into position hold at a higher altitude, which did not work. The Spirit lacked enough satellite reception to hold position at all, and evaluators said its coaxial design gave the operator no physical indication of which way the aircraft was pointing.
The Teal 2 could not be launched safely without a GPS lock, and evaluators hit that problem during familiarization flights at Fort Totten in Queens before they ever reached Hudson Yards. During the Fort Totten training, Remote ID failures kept the Teal 2 from taking off until about 10 minutes of troubleshooting cleared the error. Evaluators reported latency at short distances and a shaky, pixelated video feed.
DHS lists the Skydio X10D’s autonomous navigation modes as the reason it stayed stable in the same urban canyon, where it scored 3.7.
Links Dropped 50 Feet From Launch and Deployment Ran to 15 Minutes
The Parrot ANAFI USA GOV lost its connection as close as 50 feet from the launch point at Hudson Yards, and one loss of link ended in a flyaway. Its command and control link scored 1.8 out of 5 in the complex urban round, and its latency scored 1.7.
Evaluators there could not take off between two high-rise buildings. They reset the radio to clear connectivity errors and moved to other locations to launch.
The scale runs from 1, meaning the product met none of the evaluator’s expectations, to 5, meaning it exceeded them. A 3 means it met most of them.
In Miami, the Astro began losing video and control at about 750 feet from the controller during daytime flights at Bayfront Park. The Teal 2 had pairing and Remote ID problems in both Texas and Miami that delayed launch by up to 15 minutes, and its deployability score in Miami was 1.5. The Skydio X2D, tested only in Texas, took about two and a half minutes to recalibrate after every battery change, which evaluators called too long to get the aircraft back in the air.
In one X10D flight, an evaluator’s controller reset and hard rebooted mid-flight. The aircraft landed in place wherever it lost link, and the test team had to clear observers and obstacles while it used its own obstacle avoidance to find a flat spot.
“Deployment time was unacceptable for most first responders,” the report’s summary says. DHS adds that if a drone takes too long to deploy or costs too much, responders may go back to traditional methods instead.
Skydio X10D Scored Best, With Its Own Connection Problems
The Skydio X10D received the best rating of the six systems, 4.0 in Miami and 3.7 at Hudson Yards, which DHS attributes to fast deployment, intuitive controls and the NightSense infrared lighting kit. No other aircraft scored a 4 overall, the mark for meeting all expectations, in any round.
Half of the operators in Miami still saw discontinuities and pixelation in the X10D’s video feed when the aircraft was moving, and some evaluators at Hudson Yards lost connection below 100 feet and at short horizontal distances between high-rises. Adjusting the radio frequency by hand improved the signal, which evaluators counted as a disadvantage in an urban deployment. Redeployment took just under two minutes in Miami and up to three minutes at Hudson Yards, most of it spent waiting for the controller to reconnect after a battery swap.
The report also notes that as of October 2024, X10D systems were being delivered with one battery instead of three because of supply chain problems, with two more promised later. That is the same month China sanctioned Skydio and cut off its battery supplier.
DHS says cost was not a scored criterion but came up throughout the focus group and all three assessments. The X10D cost $28,382 as tested, the Spirit sat at the top of the range and the ANAFI USA GOV at the bottom. In 2021, an internal government memo obtained by the Financial Times put Blue sUAS at 8 to 14 times the price of the DJI drones they were meant to replace.
DHS Tested Chinese Drones Alongside and Left the Results Out
Covered nation platforms were flown alongside the Blue UAS at Hudson Yards because, the report says, response agencies still rely on them for reliability and overall performance. The results for those aircraft are available to public safety agencies on request through [email protected] and appear nowhere in the 47-page public document.
The executive summary says evaluators found large differences in capability between the covered nation drones and the Blue UAS they tested. It does not say which way the differences ran, and the report never names DJI or Autel. DHS says the assessment examined operator experience only and did not evaluate national security risk under any statute. A Blue List designation, the report adds, says nothing about whether an aircraft can do the job public safety needs done.
The test data is between 14 and 27 months old. The Teal 2 is now the aircraft the Air Force is evaluating a successor for under a $2.49 million order to Red Cat, and the X10D replaced the X2D in the urban rounds after joining the Blue List in mid-2024.
The evaluators came from 13 states and included pilots from the NYPD, FDNY, Boston Fire, Los Angeles Police, Michigan State Police, Texas DPS, the Coast Guard, and Customs and Border Protection. These are some of the largest public safety agencies in the country, and this is what their pilots said about the aircraft they are supposed to buy.
DroneXL’s Take
Volunteer fire departments and small police and rescue teams fly DJI and Autel for the reasons this report spells out: the aircraft power on and hold position, the link stays up, and the price is a fraction of the alternatives. An Arizona fire chief warned last December that lives would be lost if the DJI ban took his department’s air capability away, and Oregon’s aviation department counted 467 restricted drones across 25 states with up to $2 billion in national exposure. Now DHS’s own lab has written the technical half of that argument, with 26 of the most experienced public safety pilots in the country as the witnesses.
The report never says the Chinese drones flew better. DHS flew them at Hudson Yards because agencies still rely on them for reliability and performance, then declined to publish how they scored next to aircraft that could not hold a hover. My read is that the comparison stayed out of the report because it would have ended the argument. Any agency that requests the numbers should publish them.
The security case for Blue UAS was never a performance case, and the report concedes it in one line: a Blue designation says nothing about whether the aircraft can do the job. The supply chain case is thinner than it looks too: the report records the X10D shipping with one battery in October 2024, and Skydio’s own explanation to customers that month, covered in the story linked above, was a Chinese battery supplier Beijing had sanctioned. Blue UAS aircraft are not free of Chinese-made parts, the battery is the documented example, and motors and magnets are the components I would ask any Blue vendor about next. DroneXL’s position has been security standards any manufacturer must meet rather than bans by country of origin, and a Blue List that certifies cybersecurity while its aircraft drift over Manhattan makes that case in a single document.
The report ends by telling agencies to research the capabilities of available equipment before buying. Agencies did that research. The result is a DJI drone in the volunteer fire truck, and DHS just documented why.
Sources: Blue UAS for First Responders Comparative Assessment Report, SAVER-T-R-39, DHS S&T / NUSTL, September 2026; DHS SAVER program, Blue UAS for First Responders
DroneXL uses automated tools to support research and source retrieval. All reporting and editorial perspectives are by Haye Kesteloo. Read our editorial standards.
