A first-of-its-kind GNSS backup system provides independent navigation solutions using quantum magnetometry.
This is a double-plus development. Not only is it a solution that dramatically outperforms certain other GNSS backup approaches, but it is a prime example of a solution that leverages quantum sensing. It is compact enough, light enough, and has a power consumption low enough to be one of the first practical quantum sensors, integratable, not just for aviation and shipping, but perhaps even geomatics applications.
The Growing Challenge
Plenty has been written about the vulnerabilities of GNSS positioning, navigation, and timing (PNT) and warnings about overreliance on GNSS-only. Considering how deeply ingrained GNSS is into our daily lives, and its realized financial and social value, few are calling for a complete replacement. Instead, the wise approach has been to build layered systems. These are multi-sensor stacks where GNSS performs its primary role. Still, if there is loss of sky-view, spoofing, jamming, or interference, a backup system can be a “canary in a coal mine”, and take over primary positioning if need be.
There are many fully mature backup systems, for example, eLoran (ground-based radio signals). There are many kinds of beacons, and R&D into using “opportunistic signals”, like Wi-Fi, radio, and television, to help detect incorrect GNSS positions. There is a boom in the deployment of low-earth-orbit (LEO) satellites for communications, earth observations, and even navigation. While most LEO sats are quite small, might only stay in orbit for 5 years, and are built for a single purpose, the prospects of multi-tasking these for satellite navigation might not be feasible. Constellations of navigation-focused LEO sats are in development, notably by firms like Xona Space.
These GNSS backup systems are, just like GNSS, signal-based, and require substantial infrastructure. A highly desirable GNSS backup would be infrastructure independent, portable, and capable (at the very least) of letting you know if your GNSS is inaccurate and potentially compromised. Better still, what if a GNSS backup could take over positioning duties if needed?
Inertial navigation systems (INS) have been a reliable GNSS backup and have been used in integrated solutions for decades. For example, consider the post-processed kinematic (PPK) plus INS approach used for crewed airborne mapping, drones, mobile mapping, and backpack mapping. However, INS is drift-prone. It does not matter if the INS is tactical or strategic grade, with additional modalities; the drift precludes it from being able to reduce positioning errors for extended periods.
There are, though, cold-matter and quantum physics solutions. Quantum inertial navigation systems have been built for maritime and defense purposes. However, they require large, shielded chambers and vacuum systems to operate. Size, weight, and cost limit these to deployment on ships and submarines. The field of quantum sensing R&D has recently yielded a brilliantly practical solution, with low weight, size, and power consumption (SWaP). And it is now commercially available.
MagNav
Magnetic anomaly navigation (MagNav) is not to be confused with compass surveying, or those frustrating magnetic-oriented tilt compensation approaches for survey rovers. Those rely on the Earth’s core magnetic field and inherent limitations. MagNav is an approach that is passive, cannot be jammed or spoofed, and takes periodic positions by comparing to local measurements of magnetic anomalies of the Earth’s crust. It does this by analyzing patterns in anomaly maps while using software to remove interference from the aircraft itself.
There are a number of quantum sensing approaches that have been applied to magnetometry and RF sensing applications. For example, Rydberg atoms. A laser is fired at a cluster of atoms in a sealed box, causing the “orbits” of certain sub-atomic particles to expand far from the nucleus. These super-sensitive elements can be interrogated by other lasers to see how they have been affected by the nearby environment. One potential application for a Rydberg approach is for underground utility location. However, for MagNav, the focus had been on other approaches.
Building a Practical System
Q-CTRL is a global quantum technology company, headquartered in Sydney, Australia, focused on quantum AI-driven infrastructure software that boosts the performance of quantum computers and other quantum technologies like quantum sensors. That is a key distinction; advances in sensor technologies have made leaps and bounds – historically only on the hardware side, but now also in the deep analytical and algorithmic side, frequently AI boosted.
“Ironstone Opal is our MagNav system, about which we’ve recently published our test results, and have launched on the market,” said Michael Biercuk, CEO and founder of Q-CTRL. “The core of how it works is geophysical map matching, or DBRN, database referenced navigation technology. The version that we demonstrated in the technical manuscript and press release is based on magnetic anomaly measurements. These detect the Earth’s crustal field, as opposed to the core field, and then perform map-matching algorithms on top of the real-time sensing in order to enable position fixing in a way that corrects for INS drift. This is really a GPS backup technology that is layered on top of the inertial navigation system, a technology which is typically deployed in aircraft and maritime vehicles.”
If you’d like to drill deeper into the subject, and look at detailed test results, search for this paper: “Quantum-assured magnetic navigation achieves positioning accuracy better than a strategic-grade INS in airborne and ground-based field trials” (by M. Biercuk + 18) bit.ly/4lE19KU.
A note about magnetic anomaly fields from the paper:
“The magnetic field that is measured is composed of several parts. There is the core Earth field, which is described by a time-dependent model such as the International Geomagnetic Reference Field (IGRF) P. Alken (2022), and has a scalar magnitude of ~25,000 – 65,000 nanotesla (nT). On top of this, there are anomalies that arise from crustal geology and are stable in time. These variations are on the order of 10 nT – 100 nT over a few kilometers and are what is used for MagNav. Global anomaly maps have been produced, such as the Earth Magnetic Anomaly Grid Version 3 B. Meyer, 2017) or the World Digital Magnetic Anomaly Map (WDMAM) Choi et al., and can in principle be used for navigation. These are well-supplemented by higher-resolution maps developed by the geophysical surveying sector or defense agencies. Finally, there are time-dependent effects such as diurnal ionospheric variation (åß∂100 nT) and space weather arising predominantly from solar activity (up to 1000s of nT during solar events) Langel and Hinze (1998).”
The paper also outlined some of the challenges facing developers of MagNav systems:
• SWaP: size, weight, and power. How to make it small enough for deployment on many platforms, but without suffering from drift, noise, heading errors, or data instabilities.
• Platform noise: A vehicle or worksite will generate its own magnetic noise that interferes with the Earth’s magnetic signal; this must be mitigated.
• Sensor fusion, with other onboard sensors, like INS.
• Reduced calibration steps: A MagNav system can quickly create a platform magnetic profile, for say, a type of vehicle. A profile could be remembered, or even exist in an online database, to speed up future starts.
• Accounting for sensitivity to vehicle, payload, latitude, altitude, and sensor-location changes.
• Accounting for imperfections arising from adjusting anomaly maps for altitude and integrating realistic time-varying effects such as space weather and diurnal magnetic field variations
A first question a potential end user of a solution, especially one based on a whole new approach or technology, is: “How does it work?”.
Biercuk gave a great conceptual overview: “The sensor we’ve developed is called an ‘optically pumped magnetometer’ (OPM). There is a warm atom vapor, slightly above room temperature, and we probe that optically with laser light in order to determine the instantaneous magnetic field that’s experienced. This has a few real advantages. One, of course, is a common metric called sensitivity. That’s how small a signal we can detect in a unit time, and the magnetometer we produced is really world-class.”
Other quantum-leveraging magnetometer approaches, like a diamond magnetometer, utilize nitrogen-vacancy (NV) centers in diamonds to detect magnetic fields, and the Rydberg approaches with high sensitivity and resolution were not seen as being able to meet the performance levels Q-CTRL sought, so an OPM was the choice.
“We have a vapor of atoms, just a gas of atoms in a little glass cell,” said Biercuk. “Each atom has this feature that you reference, a degree of freedom called spin, that behaves like a little compass needle. They’re magnetically sensitive; we use lasers to align the atomic spins, and then other lasers to track their movement. This idea, leveraging the tiny, very sensitive, very stable ‘compass needles’, is the mechanism by which we perform the magnetometry and gain access to another key advantage – stability. The signal we measure comes from atomic physics. It doesn’t matter if the atoms are one degree centigrade warmer or one degree centigrade colder. It doesn’t matter over time what some electrical circuit on the side is doing; the output is stable, the resulting combination of sensitivity and stability, plus some other metrics that are relevant to navigation in particular, are what drove us to select the OPM approach.”
The compactness cannot be overstated. The sensor unit for the Ironstone Opal is about the size of a candy bar, and the control unit is a bit bigger than the size of a cell phone. The power draw is under 15W. Depending on the deployment’s specifics, there may be other components, but as it stands, this is small enough for airborne, maritime, mobile mapping systems, and even drones.
The specialization of Q-CTRL is in developing AI-powered infrastructure software that optimizes quantum hardware for real-world applications. This includes quantum computing, where the problem of errors from instability in hardware totally limits the field, and can be fixed by focusing on how software stabilizes the computers for field deployment. This also extends to the booming field of quantum sensors. When you take a magnetometer, a very sensitive detector of magnetic fields, and then you place it inside an airplane that is full of electronics and avionics, it’s made of metal, and, moving through Earth’s magnetic field, a system can be swamped by magnetic interference, typically called “platform noise”.
“We addressed that problem through software,” said Biercuk. “A totally new software denoising stack that broke with past approaches of how this problem would be addressed. While magnetic navigation is not new, people have come up with algorithmic approaches to try and perform platform denoising that were never really fit for purpose.”
Typically, this required elaborate, time-consuming, and mostly impractical calibration steps. For example, an aircraft might have to take off and fly in a square pattern to calibrate all the needed coefficients. In maritime history, ships might begin a journey by setting course between known points to calibrate their compasses. Surveyors might remember the cumbersome calibration dances they had to do with the old magnetic-oriented rover pole tilt compensation solutions.
For rapid response, or even practical civil aviation purposes, calibration steps are a non-starter.
“We built a real-time online denoising technology that learns only what it needs to on the fly,” said Biercuk. “There’s no significant pre-training. There’s no mass of AI data acquisition and processing ahead of time. It has been very effective in suppressing the platform noise and obviously allowing us to successfully perform real field tests where we outperform the best comparable GPS alternative by up to 100x.”
Results have been astounding, compared with conventional GPS backups, like INS. An excerpt from the paper:
“We present flight trials at altitudes from ground level to ~19,000 feet, testing various configurations of onboard and outboard-mounted quantum magnetometers, and comparing against a strategic-grade INS. Our MagNav solution achieves superior performance, delivering up to ~46× better (lower) positioning error than the velocity-aided INS; the best final positioning accuracy we achieve on a flight trial is 22m or 0.006% of the flight distance. Airborne trials consistently achieve at least 11× advantage over the INS across varying conditions, altitudes, and flight patterns. We demonstrate that the system can learn relevant model parameters online without special vehicle maneuvers, which provides robustness against various configuration changes (e.g., changing payload or latitude). Our trials also include the first successful MagNav performed in a ground vehicle using publicly available anomaly maps, delivering bounded positioning error ~7× lower than the INS, with both systems in strapdown configuration.”
More recent tests following the paper’s publication have extended the margin of advantage over the INS to 111X.
“We’re headquartered in Australia,” said Biercuk. “All of our major flight routes connecting Austral to the world either go to Singapore, Dubai, or similar to Europe—these are the major GPS denial areas in those routes. So, it is a major concern. There is growth in spoofing attacks, and one part of what we’re offering is early warning that there is a discrepancy between what the GPS positioning is suggesting and what the INS plus MagNav solution is telling you is your likely position. That discrepancy, beyond a certain number of standard deviations, can be displayed as a warning that you should check and make sure that everything is working right.”
I did feel compelled to ask about a trend, among a very few developers of alternate positioning technologies, that somehow seem to characterize GNSS as “obsolete”, and their solutions as being the ultimate replacement. Q-CTRL is not going that route.
“We publish our data; whenever we do something, we show the data. It is important to emphasize that we do not claim ever, and we would not claim, that this is a replacement for GPS. GPS remains state of the art and ours is a backup system with rapidly growing importance, because GPS/GNSS has become more vulnerable to attack right now. Commercial airliners suffer about 1,000 GPS-denial attacks per day along the major flight routes near the Black Sea, Northern Iraq, and Iran. Anything that connects Europe and the Middle East or Europe and Asia is subject to broad-based attack.”
While the big announcement and release of key test data were in April of 2025, Q-CTRL was well underway in fostering deployment for real-world applications.
“We’re offering eval kits right now,” said Biercuk. “And we have ongoing work with system integration partners to combine this with other technologies. We’re also excited to be delivering it directly to end users for their initial testing.”
Mag Maps
The compact sensor with refined denoising software is only as good as the magnetic anomaly reference maps. There are many publicly available magnetic anomaly maps. For example, the U.S. National Oceanic and Atmospheric Administration (NOAA) EMAG dataset and maps. The World Digital Magnetic Anomaly Map (WDMAM) was developed by an international commission. And there are other global land regional maps. However, many of these were compiled from legacy data, with mixed collection methods and technologies, and some have not been updated in two decades.
While there might be little change in most regions, to truly realize the full benefits of MagNav for expanded applications, updated maps may be necessary.
“What we see going forward is that for the core navigation applications –we focused on commercial aviation, drones, and defense operations –this technology could be limited by map availability and quality,“ said Biercuk. “In some places, we have great maps, and in other places not-so-great.”
Innovation almost always leads to further innovation. Personal navigation, like with MapQuest decades ago, led to expectations and demand for further development of Google Maps, et al. Then Streetview, and interior views, and we’re beginning to see 3D models of these publicly accessible digital worlds. Build it, and they will come. This could apply to mag maps. With mass data capture so common and persistent in ongoing capture, magnetometers could be added to sensor stacks to improve mag maps.
For many areas, static mag maps are sufficient, but in certain environments, currency of the maps can make a difference.
“In the demonstrations that we published, we also performed the world’s first successful ground-based magnetic navigation. We put this in a van, and we drove it just on a regular road, not anything special, and it worked in low-built-form environments like forests,” said Biercuk. ”It seems like it can work pretty well in areas that are heavily built up as well, though you become dominated by the anthropogenic sources: buildings, and whatnot. So, there is a critical need: if you want to do map matching in that area you must build a comprehensive and up-to-date map, so your data doesn’t get stale because someone built a new power line or built a new building. It is a much harder prospect in the urban environment, but it’s not impossible.”
Present mag maps are compiled from satellite magnetometry, airborne, mobile, and even static campaigns. Magnetometry is not an alien concept to surveyors and geospatial professionals; you’ve likely used magnetometry-based detection for underground features. A pioneer of and namesake of a popular detector producing firm, Erick Schonstedt, designed magnetometry instruments for over 400 satellites. Better magnetometers, quantum magnetometers, could beget better mag maps. These new types of sensors might find utility for a broad range of applications. Hopefully, broad enough to spur mag mapping improvement.
Geomatics applications?
While mostly confined to conflict zones, spoofing and jamming are a persistent reality. When thinking beyond defense needs, GNSS backup systems for commercial aviation and shipping are an imperative. This applies to any safety-of-life applications, for example, vehicle autonomy. Wisely, developers of vehicle autonomy employ multi-sensor stacks. In addition to GNSS, there’s usually INS, cameras, radar, and LiDAR. A compact and practical MagNav system could become part of these stacks.
Q-CTRL has grown so rapidly since its founding in 2017. There was investment from a major commercial air and space manufacturer, and a lot of attention from the maritime and defense sectors. However, when the MagNav announcement was made, and especially about the comparative testing, the wheels started turning in the minds of many folks outside of those core sectors.
I was not alone in thinking about mobile mapping. Not necessarily just for spoofing and jamming considerations, but, as the system is not as subject to drift as legacy INS, integration could yield further benefits. “We were approached by a mobile mapping organization, after our announcement, to discuss the potential for integration for incidental mapping,” said Biercuk.
For that matter, it would be practical to integrate into mapping drones, as well as backpack mobile mapping systems—any application where the GNSS could use a “second opinion”.
Could we see one on a GNSS rover? Perhaps that might be overkill, unless one is doing work in an area subject to spoofing and jamming. But perhaps not. It might have to sit in a backpack or a bulky enclosure on the pole, but it would definitely tell you if your GNSS was giving you an unrealistic position. Beyond that, the superior performance over legacy INS should be explored for other aspects of position stabilization.
Quantum sensing, which we explored in the xyHt Heights 2025 cover story, has been headed our way, though mostly on a distant horizon. This system from Q-CTRL shows us that quantum sensing is hitting much sooner than we imagined. Beep on!