59 minutes ago
16
GPS is both the backbone of modern military operations and one of its most obvious vulnerabilities. Jam the signal, spoof the coordinates, and even the most sophisticated drone becomes a very expensive paperweight. SandboxAQ thinks it has a fix, and it just proved the concept in the air.
On September 9, 2026, SandboxAQ completed what it describes as the world’s first flight test of its AQNav magnetic navigation system aboard Northrop Grumman’s Lumberjack attritable unmanned aircraft system. The milestone matters because the Lumberjack is not a research platform. It is a one-way attack drone built for contested environments where GPS is precisely the thing you cannot rely on.
How you navigate without GPS
Think of AQNav like fingerprint recognition, but for the planet. The Earth’s crust has a unique magnetic signature across every point on its surface, shaped by centuries of geological activity. AQNav reads those magnetic anomalies in real time and cross-references them against a pre-existing map, pinpointing location without ever touching a satellite signal.
The system combines quantum sensors with advanced AI algorithms to do this at speed and at scale. Quantum sensors are exquisitely sensitive to tiny fluctuations in magnetic fields, which is exactly what makes them useful here. Where a traditional magnetometer might blur the detail, a quantum sensor captures the fine-grained magnetic texture of the terrain below.
The output is what the military calls Required Navigation Performance, or RNP, a standard that defines acceptable position error tolerances for a given flight operation. SandboxAQ says AQNav has met RNP standards and has outperformed traditional inertial navigation systems, which are the current fallback when GPS is unavailable. Inertial systems work by dead-reckoning from a known starting point, accumulating small errors over time until the position estimate drifts. AQNav, by contrast, continuously corrects itself against the magnetic map.
Why the Lumberjack is the right test platform
Northrop Grumman’s Lumberjack is roughly 1.8 meters long and built around the concept of rapid reconfiguration. It is designed for swarming, strike, and intelligence, surveillance, and reconnaissance missions, and can be launched through multiple methods, each affecting its effective range. The drone is classified as attritable, meaning it is cheap enough to be considered expendable in combat.
The Lumberjack test extends AQNav’s proven track record to this class of platform. Prior evaluations were conducted on the C-17 Globemaster III, a massive strategic airlifter, alongside participation in various military exercises starting in January 2023. Demonstrating the same capability on a compact, one-way strike drone closes a significant gap in the technology’s applicability.
The flight hours behind the headline
SandboxAQ has accumulated more than 200 flight hours and over 40 sorties across multiple military aircraft since its first evaluations in 2023, supported by several Small Business Innovation Research contracts with the U.S. Air Force.
SandboxAQ is a spinout from Alphabet, Google’s parent company, and operates at the intersection of quantum technology and AI. The company has positioned AQNav as part of a broader suite of quantum sensing applications for defense, with magnetic navigation representing one of its most mature and demonstrable capabilities.
Inertial navigation, the incumbent fallback, degrades over time in a way that limits its usefulness on longer missions. Terrain-referenced navigation requires consistent elevation relief to work effectively. Magnetic navigation, by comparison, works over flat plains and open ocean with equal reliability, anywhere the magnetic map exists and sensors are sensitive enough to read it.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
English (US) ·