The navigation stack is getting bigger; is it a good thing?

For decades, it was relatively simple to describe navigation. Satellites transmitted signals, receivers processed them, and users acquired position.

Nowadays, changes to this model have become inevitable.

Autonomous systems have since become more capable and navigation environments more challenging. Thus, positioning is now built from multiple sources rather than a single dominant signal. 

GNSS remains substantial for this architecture, but inertial sensors, cameras, radar, terrestrial signals, magnetic sensing, precision timing, and even signals from communications networks can all contribute to a navigation solution. The outcome then is a navigation stack that is becoming more complicated.

That is not necessarily a problem. In fact, it may be exactly what is needed to make navigation more reliable.

Why is the navigation stack getting bigger?

GNSS is highly effective under open-sky conditions, but its signals are prone to interference, get easily blocked by structures and terrain, and become unreliable in environments where they’re most needed.

The response has since steered clear of finding one technology that can replace GNSS everywhere. Instead, today’s engineers have begun to combine different technologies that each adds a component unavailable in the others.

For example, an inertial navigation system can calculate its movement even in the absence of external radio communication. Cameras can recognize objects and their motion. Radar can supply data about the environment around it. Magnetic sensors can utilize the Earth’s magnetic field differences. Terrestrial radio signals can provide additional ranging or timing information.

Each source has its own downsides, and that is precisely why combining them is valuable.

From single-source navigation to sensor fusion

The concept is not new. Aircraft, spacecraft, ships, and military platforms have long combined inertial systems with GNSS and other sensors.

What is changing is the breadth of the inputs and the sophistication of the software combining them.

A contemporary navigation system may no longer ask simply about GPS-derived locations. Rather, it might pose certain concerns such as whether the GPS location coincides with the inertial solution, whether the camera-detected motion is compatible with the reported path, whether the radio signals confirm the same location, and whether the collected data is physically plausible.

This is where sensor fusion becomes more than a way of improving accuracy and becomes a way of evaluating trust.

The U.S. Army has been pursuing this approach as part of its Assured PNT efforts. In 2025, the Army reported fielding approximately 27,000 M-code-capable receivers in the previous fiscal year, more than 2,500 ground Assured PNT systems, and 7,000 precision guidance kits. It also said that newer systems were incorporating technologies such as vision-aided and alternative navigation alongside M-code.

The underlying principle is straightforward: if one source becomes unreliable, the navigation system should have other information available to help identify the problem and maintain an operationally useful solution.

More sensors do not mean every sensor is equal.

It is, on the other hand, tempting to think that adding more sensors somehow means creating a more robust navigation system.

But that is not the case.

Each one of them has its own errors, assumptions, and dependence on the environment. 

Inertial systems can function without GNSS, but their location estimate will eventually get skewed over time. Cameras can provide plenty of useful data in visually rich environments, but they cannot work well at night or in bad weather conditions. Magnetic navigation can exploit Earth’s magnetic field but depends on the availability and quality of magnetic signatures.

GNSS itself remains highly valuable because it can provide an absolute position reference over large areas.

The objective is now less focused on finding the “best” sensor and is geared toward understanding which source is most useful under a particular set of conditions.

The navigation system becomes a decision-maker

This changes the role of the navigation engine.

Instead of simply calculating a position, it now has to determine which measurements should influence that position more than ever. 

That means navigation systems need mechanisms for detecting inconsistent or anomalous data.

Consider a vehicle travelling through an urban environment. GNSS may indicate that the vehicle has shifted several meters. Its inertial sensors suggest continuous forward motion. A camera identifies familiar road features. A map indicates that the vehicle remains within a defined lane.

The navigation engine can compare those observations.

If one source suddenly disagrees with the others, the system has an opportunity to reduce its influence or flag the measurement for further investigation. This is especially important for spoofing. Jamming can make a GNSS receiver lose its signal. Spoofing can be more subtle because a receiver may continue producing a position solution while receiving manipulated information.

A multi-source navigation architecture creates additional opportunities to detect such inconsistencies.

What does a multi-source navigation stack look like?

The emerging architecture can be thought of as a hierarchy rather than a collection of independent sensors.

• GNSS can provide an absolute global reference.
• Inertial sensors can provide continuous motion information between external updates.
• Environmental sensors can provide information about the vehicle’s surroundings.
• Terrestrial and alternative signals can provide additional positioning or timing measurements.
• Precision clocks can preserve timing when an external reference becomes unavailable.
• Navigation software brings these measurements together, weighs their reliability, and produces the final navigation solution.

The system does not necessarily need all sources to work all the time, but enough trustworthy information to continue operating.

Why this matters for autonomous systems

The transition toward multi-source navigation becomes more and more important as autonomy increases.

A human operator can recognize that a navigation display looks suspicious. An autonomous system may not have that luxury. It must determine whether the information it receives is reliable before using it to make decisions.

That makes navigation integrity highly important.

Not only does a self-driving car, autonomous aerial device, or uncrewed maritime craft require coordinates, but also confidence that these coordinates reflect the truth.

This is why the future of autonomous navigation hinges on knowing when not to trust sensors just as much as on attaining centimeter-level accuracy.

The navigation stack will keep growing.

The expansion of the navigation stack is unlikely to stop.

Research and development are already exploring quantum sensing, signals of opportunity, vision-aided navigation, and more sophisticated inertial technologies.

The U.S. Army’s 2025 Assured PNT program, for example, specifically identified software-defined radios, vision-aided navigation, and alternative navigation as areas for next-generation capability.

At the same time, Safran itself is moving in this direction. In 2026, Safran’s LEO-PNT initiative outlined how low Earth orbit signals can be layered alongside existing GNSS and inertial technologies to deliver more precise, more resilient, and more sovereign positioning and timing services, reflecting the same shift toward integrated, multi-source architectures playing out across the industry.

These developments point toward a broader change in navigation philosophy. Hence, the question is no longer simply which technology provides the most accurate position, but about how several imperfect technologies can work together to produce a position that can be trusted.

Bigger does not mean more fragile.

A larger navigation stack may appear to introduce more points of failure.

But if designed properly, the opposite can be true.

A system that depends on one source has a single point of failure. A system that combines independent sources can continue operating when one becomes unavailable, while also using the remaining sources to identify potentially misleading information.

That is the real value of multi-source navigation.

The future navigation system may be more complicated than the one that came before it. It may require more sensors, more computing, and more sophisticated algorithms.

But that complexity serves a purpose.

The navigation stack is getting bigger because the world is becoming harder to navigate. And when the cost of being wrong is high, having more than one way to determine where you are may be one of the most important upgrades navigation can make.