US Military Counts 378 Border Drone Incursions In 30 Days, And Rising

Main Story

The scale of unmanned aerial activity along the US-Mexico border has reached a point that the Pentagon can no longer treat as a background operational nuisance. Air Force Brigadier General Brian Filler, deputy commander for operations at Joint Task Force Southern Border, told reporters at Yuma Proving Ground in late September that the military had detected 378 drone incursions in the preceding 30 days — and that the trajectory was still moving upward.

Filler described the aircraft as "largely commercial off-the-shelf available drones," most equipped with surveillance cameras and observed in correlation with cross-border movements. Army officials have characterised the typical platform as a commercial model adapted for observation of smuggling and human trafficking operations — not sophisticated battlefield hardware. Military officials have separately noted that the drones crossing the southern border are not comparable in capability to the systems seen in Ukraine or the Middle East.

The disclosure came during Falcon Peak 26.2, the fourth iteration of US Northern Command's recurring counter-UAS exercise series, which ran at Yuma Proving Ground, Arizona, from August 31 to September 25. It was the first Falcon Peak held at Yuma and the first designed explicitly around the conditions and threat profile of the southern border. The exercise brought together NORTHCOM, the Defense Innovation Unit, the Army-led Joint Interagency Task Force 401 (JIATF-401), and the Department of Homeland Security — a multi-agency integration that was itself a series first.

The urgency behind the exercise is underlined by the operational record. Since deploying laser-based defeat systems to the Rio Grande Valley on August 25, the military has downed 11 cartel-linked drones using directed-energy weapons. Broader counter-UAS operations have neutralised more than 300 drones since the start of 2026 using a combination of kinetic and non-kinetic systems, with over 100 defeated in August alone.

The next Falcon Peak iteration, already in planning, will shift the operational environment entirely — moving from open desert to a dense urban setting to stress-test low-collateral defeat requirements under the most operationally complex civilian conditions.

Technical Breakdown

Exercise Structure — Three Capability Lanes

Falcon Peak 26.2 was structured around three permanent test lanes now established at Yuma Proving Ground:

• Sensing lane: radars and electro-optical systems for drone detection, classification, and tracking.
• Electronic defeat lane: systems for jamming, spoofing, or taking over target UAS.
• Kinetic lane: ballistic systems (including automated gun platforms) and drone-on-drone intercept systems for physical takedown.

Yuma's desert environment — characterised by intense heat, dust, high winds, and glare — was treated as an active test variable rather than a logistical inconvenience. Companies reported having to adapt system configurations in real time; Allen Control Systems' team, for instance, had to account for unpredictable crosswinds affecting the flight path of interceptors launched from its Bullfrog automated gun system.

Command and Control — Anduril Lattice as Digital Backbone

For the first time in the Falcon Peak series, Anduril's Lattice autonomous C2 software platform was deployed as the exercise's digital backbone and the baseline standard against which all participating systems were evaluated. This followed JIATF-401's selection of Lattice as its enterprise tactical C2 solution under an approximately $87 million task order. All 21 commercial participants were required to feed their sensor data, radar tracks, and targeting outputs directly into the Lattice infrastructure — a strict interoperability requirement that established a common operational picture across heterogeneous sensor-to-shooter chains.

Lattice is an open-architecture platform that fuses inputs from acoustic, electro-optical, RF, and radar sensors, and applies AI-driven analytics and mesh networking to detect, track, classify, and coordinate responses to drone threats. Its design allows integration of both legacy and newly fielded systems within a single environment.

Threat Profile — Group 1 and Group 2 UAS

The incursion drones documented by Joint Task Force Southern Border are predominantly Group 1 and Group 2 commercial UAS — small multirotor platforms, consistent with widely available consumer and prosumer models such as DJI Mavic-class aircraft. Their primary utility in the observed operations is surveillance: monitoring security patrol patterns, tracking asset movements, and providing operators with situational awareness in support of illicit cross-border activity.

Participant Hardware — Key Systems on Evaluation

• Teledyne FLIR Cerberus: radar-and-camera integrated sensing system evaluated on the sensor lane.
• Allen Control Systems Bullfrog: automated machine gun system designed for kinetic drone defeat.
• Anduril Ghost-X: interceptor drone demonstrated in the kinetic lane.
• Perennial Autonomy Merops / Bumblebee / Hornet: AI-enabled interceptor and strike platforms from the holder of a $500 million Pentagon IDIQ, included as the kinetic intercept category representative.
• Harmattan AI: detection and classification system, whose team reported having to iteratively adapt to harsh desert operating conditions during the exercise.

Directed-Energy Systems — Already Operational

Separate from the Falcon Peak evaluation, laser-based defeat systems have been operationally deployed to the Rio Grande Valley since August 25. As of early October, these systems had destroyed 11 cartel-linked drones. Early reporting from the Rio Grande sector indicated a roughly 75% reduction in drone flights in that corridor following laser deployment — though Filler's border-wide figures indicate overall incursion numbers have continued to rise across the full southern border.

Low-Collateral Defeat as a Design Constraint

A recurring theme across all three capability lanes was the requirement for low-collateral defeat: the ability to neutralise a drone without producing dangerous debris fall, uncontrolled projectile trajectories, or electronic effects that could disrupt civilian communications or controlled airspace. This constraint — which does not apply in the same way to overseas combat operations — shapes both system selection and rules of engagement for border operations.

Industry Impact

For Counter-UAS Manufacturers

Falcon Peak 26.2 has effectively functioned as a structured competitive evaluation under real operational conditions, with a Pentagon contract opportunity as the explicit endpoint. The three-lane architecture creates a reproducible, standardised test environment — the lanes are permanent at Yuma — meaning vendors can expect iterative evaluation cycles rather than one-time demonstrations. The Lattice interoperability requirement sets a technical bar: any system that cannot integrate with Anduril's C2 platform is, in practice, disqualified from the Pentagon's current C-UAS procurement pipeline.

The exercise also reinforces the "Ukraine-to-border" technology transfer model. Perennial Autonomy's Merops interceptor — credited with downing more than 4,000 Russian drones over Ukraine — arrived at Falcon Peak 26.2 already holding a three-year, $500 million IDIQ with JIATF-401. The trajectory from combat-proven platform to domestic border defence contract is now an established procurement archetype that other vendors will seek to replicate.

For Platform Operators and Integrators

The shift from isolated sensor deployments to a Lattice-centred, multi-vendor, sensor-to-shooter architecture has significant implications for systems integrators. The requirement that every participating sensor feed into a common C2 picture — rather than operating as a standalone tool — raises the integration complexity and cost threshold for vendors, but also creates opportunities for middleware and data-layer companies. Integrators working with border security agencies, critical infrastructure operators, and port authorities should anticipate Lattice compatibility becoming a baseline procurement requirement across federal C-UAS contracts.

For Regulators

The next Falcon Peak iteration's urban setting signals that the regulatory challenge of counter-drone operations is moving from open-range environments toward densely populated areas where FAA airspace rules, RF interference constraints, and public safety considerations create an entirely different compliance matrix. The military's explicit focus on low-collateral defeat — accounting for debris, projectile safety, and civilian communications — foreshadows the technical standards that civilian regulators will likely need to formalise for non-military C-UAS deployments in urban areas.

For Investors

The 378-incursion figure, combined with the military's statement that the trend is rising, establishes a durable, data-backed demand signal for counter-UAS hardware and software. The Pentagon's approach — maintaining a portfolio of multiple vendors across sensing, electronic defeat, and kinetic categories, evaluated through recurring structured exercises — points toward a multi-vendor market structure rather than winner-take-all procurement. Investors with positions in early-stage C-UAS startups should watch Falcon Peak participation lists as a leading indicator of contract readiness.