The growth of spatial computing and physical AI is driving demand for more real-world data. In this context, some of the world’s leading providers of geospatial data are partnering to create something unique: Project Orbion. The companies are touting this multi-company collaboration as the next generation of spatial technology, creating a live digital twin of the earth that has accurate physics, real-time weather and more.
Leading the project is Aechelon, a maker of training and simulation systems for defense applications, which is bringing together all the various technologies into a single platform. The first major customer for Project Orbion is the U.S. Coast Guard, which will leverage the capabilities of Aechelon and the other Orbion partners into a single highly accurate, real-time solution.
What Is Project Orbion?
Project Orbion can be seen as a culmination of the work Aechelon has been doing in the defense sector for decades with its “Synthetic Reality” data, using real-world data paired with the capabilities of its new partners. First — and possibly most consequential — is the partnership with Niantic Spatial, which is providing its large geospatial model reconstruction and visualization services. Niantic’s LGM allows for the creation of highly detailed maps and can enable certain autonomous vehicles, including robots, cars and drones, to better navigate the real world with improved models and accuracy.
Niantic has also talked about pairing its Visual Positioning System data with Aechelon’s simulations systems, which should help with navigation in GPS-denied environments (for example, certain combat zones) when satellite signals are unavailable. Niantic Spatial claims that its VPS can supply ground teams with centimeter-level localization and navigation, enhancing Aechelon’s simulation accuracy.
Partner company ICEYE brings value to the Orbion initiative by providing additional data from ICEYE’s SAR satellite constellation, which delivers real-time multi-spectral 3-D processing, accelerated by GPUs. This imagery is unaffected by night, weather or smoke, making it extremely valuable for partners like the Coast Guard that benefit from high-resolution data, for example in search-and-rescue missions during major catastrophes such as hurricanes and wildfires.
Another partner, Blacksky, provides additional information with its own satellite constellation and imagery as well as AI-accelerated intelligence about aircraft, ships and other vehicles. Blacksky has its own platform called Spectra that can further enhance the real-time accuracy of the Synthetic Reality global digital twin.
Finally, partner Distance Technologies makes light field 3-D displays that allow aircraft, ships and other vehicles to see through the “fog of war” — or any kind of bad visibility stemming from weather or other circumstances. This technology is still in development, but I know that the company already has some early defense customers, and I believe these displays will find their way into helicopters and planes that need to have an accurate depiction of the real world regardless of visibility conditions.
Project Orbion’s Future
Aechelon already has one of the world’s most effective simulators for the defense industry. The Project Orbion partnership with the Coast Guard truly has the potential to save many lives by enabling better training and situational awareness during catastrophes. That said, there’s also an entirely different angle of application for robotics, autonomous vehicles or anything that would benefit from world-scale maps that are accurate and updated in real time. While this isn’t necessarily the best data source for training an autonomous car to navigate highway traffic, it could absolutely be used to train AI models to accurately steer autonomous boats, autonomous mining equipment or autonomous drones.
The potential physical AI applications are numerous, and I believe the Orbion simulator could be used to make models more accurate and capable based on real-world data — which is ultimately the most valuable when trying to perfect a physical AI model. Also, when you consider how small some of these physical AI models have to be to run locally, you realize that model quality is of the utmost importance, and I can’t think of a more valuable data set than what Aechelon has put together with Project Orbion.
Moor Insights & Strategy provides or has provided paid services to technology companies, like all tech industry research and analyst firms. These services include research, analysis, advising, consulting, benchmarking, acquisition matchmaking and video and speaking sponsorships. Moor Insights & Strategy does not have a paid business relationship with any company mentioned in this article.
