Imroatul Nasiyah
Most of us rarely question the accuracy of the GPS dot that shows our location on a map, a ubiquitous digital compass guiding our daily commutes and adventurous explorations. Yet, when venturing into the labyrinthine embrace of a new city, this seemingly infallible technology can falter dramatically, leaving us disconcertingly adrift. It’s a common, frustrating experience: you’re walking at a steady pace along a clearly defined sidewalk, but on your phone’s map, your digital avatar appears to be performing erratic leaps and bounds, jumping from one spot to another, a phantom dancer in the urban landscape. This jarring disconnect between our perceived movement and the technology’s representation is not a mere glitch; it’s a fundamental challenge posed by the very fabric of urban environments.
"Cities are brutal for satellite navigation," explains Ardeshir Mohamadi, a doctoral fellow at the Norwegian University of Science and Technology (NTNU). Mohamadi and his team are at the forefront of a critical research endeavor: developing more precise and affordable GPS receivers, the kind that power our smartphones and fitness watches, without the reliance on costly external correction services. This pursuit of heightened accuracy is not just about convenient navigation; it holds paramount importance for the burgeoning field of autonomous vehicles, where self-driving cars depend on an unwavering understanding of their precise location to navigate safely and efficiently.
The culprit behind these urban navigation woes is a phenomenon known as "urban canyons." This term vividly describes the challenging environments created by dense concentrations of tall buildings, constructed from materials like glass and concrete. These structures act as formidable barriers and deceptive mirrors for satellite signals. "In cities, glass and concrete make satellite signals bounce back and forth," Mohamadi elaborates. "Tall buildings block the view, and what works perfectly on an open motorway is not so good when you enter a built-up area."
The fundamental principle of GPS relies on a receiver calculating its distance from a constellation of satellites orbiting the Earth. Each satellite broadcasts signals containing its precise location and the exact time of transmission. By triangulating signals from at least four satellites, a GPS receiver can pinpoint its position. However, in an urban canyon, these signals take a circuitous route. When a satellite signal encounters a building, it can reflect off its surface, a process akin to an echo. This reflected signal, having traveled a longer path, arrives at the receiver later than the direct signal. This delay introduces a significant error into the distance calculation, leading to an inaccurate positional fix. Imagine being at the bottom of a deep gorge; direct sunlight is scarce, and any light you perceive has likely bounced off the canyon walls multiple times. Similarly, GPS signals in urban canyons are often a collection of these reflected, delayed, and distorted echoes, rather than a clear, direct transmission.
This distortion is not merely an inconvenience; for autonomous vehicles, it can be the difference between confident, reliable operation and hesitant, unpredictable behavior. "For autonomous vehicles, this makes the difference between confident, safe behavior and hesitant, unreliable driving," Mohamadi emphasizes. "That is why we developed SmartNav, a type of positioning technology designed for ‘urban canyons’."
Beyond the issue of reflected signals, even the direct satellite signals that manage to penetrate the urban labyrinth often lack the necessary precision for critical applications. Standard GPS receivers, particularly those in consumer electronics, are designed for general-purpose location services, not for the centimeter-level accuracy required by self-driving cars.
To tackle this multifaceted problem, Mohamadi and his team have ingeniously combined several different technologies, weaving them together into a sophisticated computer program designed for integration into autonomous vehicle navigation systems. Their approach involves correcting the compromised satellite signals by leveraging a combination of existing and emerging technologies.
To understand their solution fully, it’s essential to briefly revisit how GPS operates. The Global Positioning System (GPS) consists of a network of satellites that continuously transmit radio wave signals. A GPS receiver on Earth picks up these signals. The signal itself contains crucial data: the satellite’s orbital information (its position in space) and the precise time the signal was sent. This information is akin to a timestamped text message from space. When a receiver can lock onto signals from at least four satellites, it can calculate its distance from each by measuring the time it takes for the signal to arrive. With these distances, it can then perform a process called trilateration (or more accurately, multilateration) to determine its own three-dimensional coordinates on Earth.
The core of the urban GPS problem lies in the distortion of that "timestamped text message." When signals bounce off buildings, the encoded time information can become corrupted, or the signal itself might be so weakened and delayed that it’s unusable. One of the initial avenues of research explored by the NTNU team was to discard the timing code altogether. Instead, they focused on extracting information directly from the radio wave itself.
This alternative approach centers on analyzing the "carrier phase" of the radio wave. The carrier phase describes the wave’s oscillation, specifically whether it’s traveling upwards or downwards as it reaches the receiver. "Using only the carrier phase can provide very high accuracy, but it takes time, which is not very practical when the receiver is moving," Mohamadi explains. The limitation here is that precise carrier phase measurements require the receiver to remain stationary for an extended period – not just a microsecond, but potentially several minutes – to allow for the signal to stabilize and be accurately analyzed. This is clearly not feasible for a moving vehicle.
Recognizing the limitations of relying solely on the carrier phase, the researchers turned to other established and developing methods for improving GPS accuracy. One such method is Real-Time Kinematics (RTK). RTK systems employ ground-based reference stations that have precisely known locations. These base stations receive the same satellite signals as the mobile receiver and can calculate the errors in those signals. They then transmit correction data to the mobile receiver, allowing it to correct its own position with high accuracy, often down to a few centimeters. However, RTK systems typically require a dense network of these expensive base stations and are therefore primarily used by professional surveyors and in specialized industrial applications.
A more advanced and increasingly accessible approach is Precise Point Positioning – Real-Time Kinematic (PPP-RTK). This technology combines the benefits of precise point positioning (which uses satellite orbit and clock data for initial corrections) with the real-time kinematic corrections. The European Galileo satellite navigation system, a competitor and complement to GPS, has been instrumental in advancing PPP-RTK by broadcasting its correction data free of charge. This significantly reduces the need for local base stations and makes high-precision positioning more broadly available.
But even with these advancements, the urban canyon problem persisted. This is where a significant breakthrough emerged from an unexpected source: Google. While the NTNU researchers were diligently refining their algorithms, Google was simultaneously developing its own sophisticated approach to urban navigation.
Google’s innovation lies in its extensive use of 3D building models. Imagine planning a trip to a city like London. With Google Maps on your device, you can not only see your route but also zoom in to study the streetscape, observe the facades of buildings, and even gauge their height. Google has meticulously created these detailed 3D models for nearly 4,000 cities worldwide. They are now leveraging these models to predict and compensate for the complex signal reflections that occur in urban canyons.
"They combine data from sensors, Wi-Fi, mobile networks and 3D building models to produce smooth position estimates that can withstand errors caused by reflections," Mohamadi states, highlighting the synergistic power of this approach. By understanding the geometry of the urban environment and the likely paths of reflected signals, Google’s algorithms can more accurately filter out erroneous data and provide a more stable and reliable position estimate. This directly addresses the frustrating "wrong-side-of-the-street" problem, where a GPS signal might incorrectly place you across the road from your actual location.
The NTNU researchers then took their own sophisticated algorithms and integrated them with these various correction systems, including the data provided by Google’s advanced mapping capabilities and the free PPP-RTK corrections from Galileo. The result of this multidisciplinary collaboration is a powerful system that, when tested in the challenging urban streets of Trondheim, achieved an astonishing level of accuracy. They reported achieving positional accuracy better than ten centimeters 90 percent of the time.
This level of precision is a game-changer for urban navigation, especially for autonomous vehicles. It provides a level of reliability that was previously unattainable in dense city environments. Furthermore, the integration of PPP-RTK and Google’s data makes this high-precision technology accessible to a much wider audience. "PPP-RTK reduces the need for dense networks of local base stations and expensive subscriptions, enabling cheap, large-scale implementation on mass-market receivers," Mohamadi concludes. This signifies a future where not only self-driving cars but also everyday consumers can benefit from near-centimeter accuracy, transforming how we navigate and interact with our urban surroundings, making cities less of a "brutal" environment for our digital navigators and more of a seamless, interconnected space.
