For anyone developing products and services that rely on positioning, navigation and timing (PNT) data, 2024 promises to be an exciting year. With ongoing advances in artificial intelligence, edge computing and the emerging availability of low-Earth-orbit (LEO) satellites, the next 12 months will see new ways emerging to obtain high-precision location and timing information. More broadly, there will also be opportunities to speed up the time to market of new products and services.
In this article, we will look at how advancements in AI, edge computing and LEO satellites are shaping PNT opportunities in 2024.
AI: new ways to achieve high-precision positioning
Like all parts of the technology industry, PNT is being shaped by advances in AI, including generative AI (genAI) and machine learning (ML). Although the high compute requirements of AI and ML mean their use in low-power devices like consumer wearables or wireless tracking tags will remain challenging, it’s a different story in industrial and automotive applications.
The trend in the automotive sector toward compute centralization, for example, means many vehicles now have powerful on-board processors. These bring new opportunities to use AI and ML to augment data from global navigation satellite systems (GNSS) and other in-vehicle PNT sensors, such as LiDAR, cameras and inertial measurement units (IMUs).
This will offer new ways to overcome some of the drawbacks of conventional sensors, such as IMU drift. By doing so, AI and ML will offer manufacturers new ways to move toward the centimeter-level positioning required for higher levels of vehicle autonomy, including in tunnels, parking garages and other locations where obtaining an accurate and reliable GNSS-only position reading is challenging.
Elsewhere, we expect to see genAI dramatically accelerating the development of new positioning and timing capabilities in 2024. The word “transformative” is overused, but in this context, it’s entirely appropriate: GenAI chatbots and copilots have enabled developers to complete tasks in minutes that once took days.
Many more organizations will adopt these AI-enabled development tools during 2024. The resulting increase in productivity will mean everyone in the industry will be able to speed up the time to market for new PNT services and capabilities, and it will particularly benefit those with smaller research and development teams, such as startups and small and medium-sized enterprises.
Edge computing: enables new positioning applications
Edge computing places powerful processing capabilities close to the source of positioning or timing data—in this context, the receiver. Doing so enables device makers to perform sophisticated tasks without the need to send data to the cloud, which may not be possible or practical in all applications or locations.
Using a vehicle’s on-board processing capability for AI and ML to augment GNSS sensor data, as we touched on above, is a great example of edge computing in practice. We foresee many more uses of edge computing in PNT applications emerging in 2024.
In the consumer space, for example, some wearables can’t currently detect very small positional movements, such as a golfer who is walking a short distance between putting attempts while wearing a tracker watch. Device makers could address this by offloading the positional calculations from the watch to the wearer’s smartphone. In addition to unlocking much more computing power, this can enable the application to access other sensor data to give a higher level of positional accuracy and, by extension, offer a more attractive or lucrative service to customers.
Industrial applications stand to benefit as well. As an example, companies that are providing logistics tracking services that use GNSS-enabled tags could start offering customers richer insights or extend tag battery life by using short-range communications to offload processing and communication of the data to edge infrastructure. Such edge infrastructure could be located on the higher-level infrastructure, serving multiple endpoints—e.g., on a ship or at a container facility. Here, additional processing of the data could be carried out, that is potentially power-hungry, before sending data onward to the cloud.
LEO satellites: emerging to complement MEO GNSS
2024 will see continued enhancements being made to conventional GNSS technology, where the satellites operate in medium-Earth orbit (MEO). Among these improvements, we’ll see even more satellites broadcasting modernized L5 signals, which help overcome issues like jamming, ionospheric delay and multipath when used in a dual-band L1/L5 setup.
In parallel, expect to hear about lots of developments in LEO positioning and timing capabilities. LEO will complement MEO-based GNSS by offering distinct characteristics that will be well-suited to selected applications or that help overcome certain limitations of MEO GNSS. In particular, LEO satellites will offer stronger signals (due to satellites being closer to Earth) and greater signal diversity (because satellites are in view for less time).
While we’re a good few years away from LEO constellations being widely used for PNT, 2024 should be the time you start ideating around how your customers could benefit from the new and enhanced positioning and timing opportunities it will ultimately provide.
Early uses of LEO positioning are likely to include the opportunity to offer seamless indoor-outdoor positioning and timing, thanks to the LEO signals’ ability to penetrate into buildings. Imagine how this could enhance logistics tracking, where assets pass through a variety of indoor and outdoor environments on their journey but need to be tracked throughout. MEO-based GNSS can’t cover this whole journey on its own. LEO-based positioning could fill the gaps.
Data center operations will likely be another area to benefit, by using LEO timing signals to keep networks in sync. Doing so should bring opportunities for network architects to simplify this critical but often complex part of running data centers while also boosting resilience. This is because LEO-based timing promises to reduce or eliminate the need for components like external GNSS antennas and instead operate with LEO antennas directly attached to the network components, where they’re easier to access and maintain.
One of the reasons we’re a little way off using LEO for PNT is that there are currently no governmental positioning constellations in this orbit. Public ownership of the MEO constellations, and the fact that GNSS signals have been available free of charge, were key drivers of GNSS adoption and the development of what’s become a comprehensive device ecosystem. This is why we’re pleased to see governments beginning to develop LEO PNT capabilities. The European Space Agency went to tender in 2023 to develop a test LEO PNT constellation, for example.
In parallel, it will be fascinating to watch how the private sector offering in this space evolves during 2024. There’s plenty of activity going on, but putting things in space is expensive, and not all of those currently planning LEO constellations will necessarily secure the funding they need to succeed in the medium to long term.
AI, edge computing and LEO for PNT are long-term trends that will continue to shape this space for many years. However, it’s clear that there are plenty of steps product makers can take in 2024 to begin benefiting from developments in these areas over the coming months. In addition, by laying the groundwork this year to leverage some of the developments coming down the line, they’ll be putting themselves in a strong position for the medium to long term.
About the author
Dr. Markus Uster heads the Product Center Positioning at u-blox. In this role, he is responsible for formulating and implementing the strategy as well as the corresponding product portfolio and roadmap of the Product Center. He also spearheads technology development efforts, provides support for cross-departmental activities, and ensures the overall commercial success of the Product Center. His main interests include adjusting the organization to be “business agile” and enabling further growth.
Markus came to u-blox with over 20 years of experience in R&D and managerial positions at well-known companies such as Landis+Gyr, X-Rite and Mettler-Toledo. He received his master’s and Ph.D. degrees in electrical engineering from the Swiss Federal School of Technology (ETH), Zürich, Switzerland.
Learn more about u-blox