When the Department of Homeland Security's own test laboratory flew a set of Blue UAS-listed drones through the glass-and-steel canyons of Manhattan's Hudson Yards, three of the five platforms were judged unsafe to fly. The reason was basic: they could not hold a position without drifting. The result comes from the "Blue UAS for First Responders Comparative Assessment Report" (SAVER-T-R-39), published in September 2026 by the National Urban Security Technology Laboratory (NUSTL) for the DHS Science and Technology Directorate.

The report has since become ammunition in a policy fight. The Drone Service Providers Alliance (DSPA) used it on September 27 to press Congress, arguing that Blue UAS status does not equal mission capability. Here is what the document says, and what it does not.

How the Assessment Was Run

NUSTL's report was prepared by Kris Dooley, Daniel Kadyrov, Blaise Linn and Mitchell Roberts. It draws on three field assessments in progressively harder environments:

  • College Station, Texas (rural), June 2024
  • Miami (urban), November 2024
  • New York City, Hudson Yards (complex urban), July 2025

Twenty-six experienced remote pilots served as evaluators, scoring the aircraft on a 1-to-5 scale. The DSPA describes the full document as 47 pages.

The Scores and Prices

The report lists overall scores and prices as assessed. Not every aircraft was scored in every environment.

  • Ascent AeroSystems Spirit: 3.1 urban, 3.3 rural day, 3.2 rural night, no score in complex urban conditions; $56,195
  • Freefly Astro: 3.5 urban; $47,018
  • Parrot ANAFI USA GOV: 3.1 urban, 3.3 rural day, 2.9 rural night, 2.4 complex urban; $13,964
  • Skydio X2D: 3.1 in both rural day and night; $21,889
  • Skydio X10D (with VT300-Z payload and NightSense): 4.0 urban, 3.7 complex urban; $28,382
  • Teal 2: 2.7 urban, 2.8 rural day, 3.2 rural night; $15,073

The X10D was the highest scorer in the tables, and the DSPA's Vic Moss names it as the standout. The tested systems span $13,964 to $56,195, and the report identifies cost itself as a barrier for public safety agencies.

What Happened at Hudson Yards

A footnote in the report says some platforms were "deemed unsafe to fly during the complex urban test due to their inability to maintain a position hold without drifting." The report cites RF interference, GNSS sky obscuration and multipath as contributing conditions. According to DroneXL's reading, the Ascent Spirit, Freefly Astro and Teal 2 were excluded from scoring at Hudson Yards for that reason, leaving only the Parrot ANAFI USA GOV and the Skydio X10D rated in the complex urban environment. Of those two, the Skydio scored 3.7 and the Parrot 2.4.

Four Findings

DHS summarizes the results in four takeaways:

  1. GPS is unreliable in dense urban areas. Positioning performance degraded where buildings block or reflect satellite signals.
  2. The link between ground control station and aircraft was inconsistent. Blue UAS platforms suffered disconnects "as near as 50 feet from the launch point."
  3. Deployment times were unacceptable. Causes included GCS disconnects, GPS lock and Remote ID failures, and calibration delays. The Teal 2's GCS pairing problems caused deployment delays of up to 15 minutes in some instances.
  4. Cost is a barrier to adoption by public safety agencies.

The Missing Comparison

The report's most contested feature is what it leaves out. Covered-nation drones (systems from countries restricted under federal law, which the DSPA reads as DJI and Autel) were flown alongside the Blue UAS at Hudson Yards, and the report says evaluators found significant differences between covered-nation UAS and select Blue UAS. Those results are not in the public document. The report states they are available to public safety agencies upon request. Readers therefore cannot see how the approved aircraft compare with the products many agencies already fly.

The Policy Fight

On September 27 the DSPA published a post by CEO Vic Moss, which the group acknowledges strays outside its charter. It calls the report evidence that Blue UAS status does not equal mission capability, and it hosts a copy of the report. The post urges readers to contact their elected officials and asks Congress to direct DHS, or another appropriate federal security agency, to run a cybersecurity assessment of drones operating in US airspace and to set country-agnostic standards.

The backdrop, per DroneXL: Section 1709 of the FY2025 NDAA, enacted in December 2024, required an audit of DJI and Autel within one year, a deadline of December 23, 2025, and no agency ran it. The FCC added foreign-made drones and critical components to its Covered List on December 22, 2025.

Why It Matters

Blue UAS listing is a security and supply-chain vetting mark. It does not certify that an aircraft works in the environment a first responder actually faces. NUSTL's data put numbers on that gap: in the hardest test, three of the five aircraft flown could not hold station, and link losses occurred at distances that would matter in any city response. For agencies being pushed away from covered-nation platforms, that raises a procurement question with real operational stakes.

Caveats apply. The results reflect specific aircraft configurations at specific dates, and the report's footnote ties the Hudson Yards position-hold problems to RF interference, GNSS sky obscuration and GNSS multipath. The withheld covered-nation data also means the public document cannot show whether approved drones are worse than the alternatives, only how the Blue UAS scored against pilots' expectations. Agencies that want that comparison must request it from NUSTL.

Sources