Crowded cities with stacked road systems and reflective architecture may offer impressive skylines, but for GNSS receivers, they create some of the harshest conditions on Earth. For technologies that depend on stable, trustworthy positioning, real-world testing in these challenging environments is essential. Here, Jez Ellis-Gray, product manager at Focal Point Positioning, a provider of GNSS positioning software, examines what recent field deployments reveal about the future of reliable GNSS.
Urban environments present unique constraints that no laboratory or simulation can perfectly replicate. A lab test may miss the thousands of variables that influence signal behaviour in a living, breathing city or a dense forest road. This matters most for automotive applications, where positioning must remain stable and trustworthy to support driver assistance and higher levels of automation. That is why we conducted field trials across Germany and Japan, evaluating FocalPoint’s S-GNSS Auto software running on STMicroelectronics’ Teseo GNSS receivers in challenging real-world conditions.
This testing demonstrated that GNSS performance in the real world is often determined not by peak accuracy under ideal conditions, but by the system’s reliability when satellite signals are distorted, reflected or partially obstructed.
This distinction — between accuracy and reliability — is becoming increasingly important for sectors where positioning plays a safety-critical role, including automotive.
Understanding the complexity of real environments
Germany’s combination of modern architecture and medieval street layouts made it a good place to test GNSS upgrades against standard technology. A city like Frankfurt offers a nice mix of glass facades, narrow streets and tall buildings – conditions that tend to create multipath interference.
During our recent field testing, conventional GNSS receivers frequently suffered from severe degradation of position accuracy when compared to a state-of-the-art ground truthing system. The standard receiver positions will often drift away from the travelled path, often through buildings or even onto parallel roads.
As accuracy deteriorates, the receiver can usually tell that the input information is poor, and output a warning for a larger estimate of error. This is a useful warning flag for AVs and allows for safe handover back to the human driver. However, in some cases, the reflected signals cause the miscalculation to be assumed correct. This “confident but wrong” GNSS is a much greater threat to autonomous driving, as the vehicle may make a dangerous decision based on this false information.
In contrast, the S-GNSS Auto enhanced receiver was able to maintain lane-level accuracy far longer, even in areas where intense reflections would normally overwhelm the satellite data. In multipath-heavy environments, S-GNSS on Teseo receivers showed an accuracy improvement of up to 4x. These findings reinforced our belief that, as automation increases, consistency and reliability will be more valuable than peak accuracy.
A navigation system that performs well on open motorways but struggles on urban roads will not scale safely to higher levels of vehicular automation. This is particularly relevant as the industry transitions from Level 2 to Level 3 autonomy, marking the point at which a vehicle takes full responsibility for the driving task in defined conditions, allowing the driver to disengage temporarily while the system manages safety-critical decisions.
This shift is expected to unlock significant commercial value. A 2023 report by McKinsey predicted that advanced driver assistance and autonomous driving features could generate between $300 billion and $400 billion in annual revenue by 2035, driven by software services and subscription-based functionality that depend heavily on reliable positioning. Unlocking this potential will depend on the next generation of vehicles having robust positioning systems, as users are unlikely to pay ongoing subscriptions to systems that repeatedly require human intervention or where safety concerns linger.
Japan: one of the world’s most challenging environments
If Germany represents a demanding test bed, Japan pushes GNSS to the extreme. Tokyo offers some of the toughest conditions anywhere in the world due to its towering buildings, multilevel road networks and narrow corridors that create intense multipath environments, so it was a natural choice for our next field test.
The results showed that in particularly dense districts such as Shinjuku, standard GNSS receivers often struggled to maintain a coherent position solution. Reflections from glass towers, elevated highways and rail lines produced non-line-of-sight signals that overwhelmed conventional algorithms.
However, receivers equipped with S-GNSS’s advanced signal-processing techniques demonstrated significantly improved performance. These upgraded devices maintained a stable positioning where traditional systems faltered, avoiding errors that would cause an automated system to disengage or provide dangerously erroneous positions
This improved reliability has direct implications for safety and user experience, which vehicle OEMs will no doubt welcome. In driverless vehicles, GNSS problems that trigger sudden driver handovers or interruptions to hands-free modes, are likely to erode trust and reduce the likelihood of subscription renewals, as the end user will judge the product less by its peak performance and more by its dependability in everyday situations.
Field testing and the future of positioning technology
As cities evolve, buildings grow taller and mobility systems become more congested, the challenges facing GNSS will only increase. As such, automotive OEMs are rightly starting to demand real world results, not just in ideal conditions (static, open sky) but in the worst conditions.
Manufacturers increasingly recognise that positioning is now a foundational technology that underpins safety, automation and customer experience. Investments in more reliable GNSS systems are therefore not marginal enhancements but essential enablers of future services. For companies developing navigation and sensing technology, real-world testing offers a unique opportunity to understand how systems react to chaotic, imperfect environments. It provides granular insight into where and why positioning fails, and how these software-based enhancements can bridge the gap. By validating these solutions in the world’s toughest GNSS environments, developers can offer manufacturers greater confidence in deploying advanced features across global markets.
You can request an evaluation kit here or download the full results report of our latest testing here.
