I'm not old enough to have navigated the lighted airway system, although traces of that earlier infrastructure were still around when I learned to fly. My instrument training in the mid-1980s belonged firmly to the radio-navigation era.
Most of that training was built around VORs. We learned to intercept and track radials, determine position from crossing radials, and hold at VOR intersections. Sometimes that meant doing it with only one VOR receiver: tune one station, establish one line of position, retune the receiver to the second station to identify the intersection, then return to the first frequency to continue navigating. It demanded a fairly active mental picture of where the airplane was.
NDB navigation was part of the same training environment. An NDB transmitted a low- or medium-frequency signal in all directions and the airplane's ADF indicated the relative direction to the station. It gave you bearing information, not distance, and tracking it meant accounting for wind rather than simply keeping the needle pointed at the beacon.
NDBs were not remnants of the original lighted-airway navigation system in a technical sense. The airway beacons and radio navigation systems were different technologies developed in different periods, although some later radio facilities could share sites where aviation infrastructure already existed. What connected them for me was the training philosophy: navigation depended on knowing how to extract a position or course from relatively simple external references.
By the time I was learning those skills, long-range radio navigation was already much more sophisticated.
LORAN-C compared the arrival times of precisely timed pulses from chains of ground transmitters. Each time difference established a hyperbolic line of position; intersecting measurements produced a geographic fix. Omega extended long-range radio navigation globally, using synchronized very-low-frequency transmitters and phase comparisons rather than LORAN's pulse timing.
Both were important transitional technologies. They were still dependent on external radio transmissions, but the airplane could now determine a position over distances far beyond an ordinary VOR or NDB network.
My first actual experience with long-range overwater navigation took me in another direction entirely: the Delco Carousel IV inertial navigation system.
The Carousel didn't need to receive anything from the ground. Once aligned and given an accurate starting position, its gyroscopes and accelerometers measured the aircraft's movement, and the system continuously integrated that motion into a calculated position. We entered latitude-and-longitude waypoints into the system and could navigate across an ocean without remaining within range of an external navigation transmitter.
That solved the coverage problem in a very different way. LORAN and Omega extended the reach of the external reference. INS removed the need for one.
But inertial navigation created its own problem: drift. Small errors in gyros, accelerometers and initial alignment accumulated as the system repeatedly integrated acceleration into velocity and velocity into position.
One of the less obvious concepts behind systems such as the Carousel was Schuler tuning. A terrestrial inertial system must continually maintain its concept of local vertical while moving over the curved surface of the Earth. Properly mechanized, small horizontal-reference errors don't simply grow without bound; they oscillate with an approximately 84.4-minute period. That doesn't eliminate inertial position error, but it makes a practical Earth-referenced inertial system possible.
The Litton LTN-92 represented another substantial step. It was a strapdown inertial system using ring-laser gyros rather than the older generation of mechanical gyros. Counter-propagating laser beams travel around a closed optical path, and rotation produces a measurable difference through the Sagnac effect.
For the operator, that meant a more accurate and reliable inertial reference. The LTN-92 could also accept external position updates, so the distinction between an autonomous inertial system and an externally corrected navigation system was already beginning to disappear.
The Boeing 747-400 pushed that integration much further. The flight management computers could bring together inertial position, radio-navigation information and, as GPS became available, satellite position. The crew was no longer managing a collection of largely independent navigation systems. The airplane was managing multiple sources and producing a best estimate of position.
The question had changed. It was no longer simply, How do we navigate when we're beyond the range of ground stations? It had become, How do we continuously determine the best position from several independent sources?
That architecture became progressively more capable in the 777 and 787. By the 787, GPS and inertial information can be combined into hybrid GPS-inertial positions. If GPS becomes unavailable, the flight management system can use inertial position corrected by DME/DME radio fixes, then other combinations of terrestrial radio and inertial information, and ultimately the inertial solution by itself.
GPS therefore did not make inertial and terrestrial navigation obsolete. It became the most accurate member of an integrated navigation architecture.
For much of my career, the accuracy and availability of GPS made that arrangement seem close to ideal. What has changed is our understanding of its vulnerability.
A GPS receiver is listening to extremely weak satellite signals. Those signals can be overwhelmed by jamming. Spoofing is potentially more insidious because the receiver can be presented with a false but apparently credible position.
After decades of improving navigation partly by adding better external references, the problem again becomes familiar: what can the airplane use when it cannot trust the signal coming from outside?
One emerging answer is quantum navigation, particularly magnetic-anomaly navigation.
The word quantum can make the system sound more mysterious than it is. The navigation principle itself is straightforward. A quantum magnetometer uses atomic-scale effects to measure extremely small changes in magnetic field strength with much greater sensitivity and stability than conventional sensors. Those measurements reveal subtle variations in Earth's magnetic field produced partly by magnetic minerals in the crust.
Those variations can be mapped. In flight, the system measures the local magnetic field, removes as much aircraft-generated and environmental interference as possible, and compares the remaining pattern with a stored geomagnetic map. The resulting match provides a position observation, which can then be blended with the airplane's existing inertial solution.
In that sense it resembles the architecture I've already watched develop. The quantum system is not necessarily the navigator by itself. It becomes another independent sensor feeding the navigation solution.
That distinction matters because the primary value is not that a quantum sensor somehow produces a perfect latitude and longitude continuously. Its value is that periodic geophysical position fixes can bound the error of an inertial system. Instead of allowing an INS position to drift indefinitely after losing GPS, the magnetic-map solution can pull it back toward a known geographic reference.
Airbus has been developing magnetic quantum navigation for aviation, while Q-CTRL's Ironstone Opal combines quantum magnetometers or gravimeters with map matching and an existing INS or IMU. In both cases, the practical architecture is familiar: an additional independent sensor is fused with the navigation system already aboard the aircraft rather than replacing everything that came before it.
The failure modes are different from GPS, not nonexistent.
The first problem is the map. Magnetic navigation works best where the Earth's magnetic field contains enough geographically distinctive structure to produce an unambiguous match. Areas with weak gradients or similar magnetic patterns can provide less position information. The quality and resolution of the stored map therefore matter.
The second problem is the airplane itself. Electric motors, wiring, avionics, ferrous structure and changing electrical loads all produce magnetic fields. Separating those effects from the Earth's signal is a major part of the engineering problem. Atmospheric and geomagnetic variations also have to be distinguished from the relatively stable geographic signature the system is trying to recognize.
There are algorithmic limits as well. A poor initial inertial estimate, inadequate map data, sensor noise or an ambiguous magnetic signature can reduce the quality of a position fix even though nobody is actively jamming the system. Quantum navigation is therefore resistant to the conventional failure modes of satellite navigation, but it is not infallible.
That may make it a natural fit with the navigation philosophy already used on modern airplanes. Instead of betting everything on one perfect source, the aircraft compares independent sources whose errors fail in different ways.
My own progression started with tuning VORs, tracking radials and holding at intersections with one receiver. NDBs required another kind of mental picture. LORAN and Omega showed how distant transmitters could produce a geographic position. The Carousel demonstrated that an airplane could navigate across an ocean without receiving any external signal at all. Ring-laser inertial systems made that self-contained solution better. The 747-400, 777 and 787 increasingly combined inertial, terrestrial and satellite information into a single navigation solution.
Now GPS jamming and spoofing have reminded us that extraordinary accuracy is not the same thing as independence.
Quantum navigation may add that independence back into the system—not by abandoning what came before, but by giving the airplane one more fundamentally different way to answer the same question I've been dealing with since instrument training:
Where are we?