GEODNET hits $8M ARR

# GEODNET hits $8M ARR A robot can already tell you it is on Main Street. The hard part is telling whether it is in the right lane, on the shoulder, or on the sidewalk, because ordinary satellite navigation is often off by several meters. (docs.geodnet.com) GEODNET says it has built a business around fixing that last few meters of error. The company’s website currently shows about 21,237 reference stations in 159 countries and about $9.37 million in annual recurring revenue, while the story circulating online frames the milestone as roughly 22,000 stations across 150 countries and $8 million in annual recurring revenue. (geodnet.com) The underlying trick is called real-time kinematic positioning, or RTK. Instead of trusting a satellite signal by itself, RTK compares that signal with measurements from a fixed ground station whose exact location is already known. (gpsworld.com) That fixed station acts like a friend standing on a surveyed spot and telling your device how far the satellite reading is drifting. GEODNET says each station can send correction data to nearby devices within roughly 20 to 40 kilometers. (gpsworld.com; geodnet.com) Once those corrections are applied, the error can shrink from meters to centimeters. GEODNET says its network delivers about a 100-fold improvement over ordinary Global Positioning System positioning and advertises 2-centimeter precision on its coverage pages. (geodnet.com; rtk.geodnet.com) That level of accuracy is useful anywhere a machine has to move through the physical world without scraping, drifting, or guessing. GEODNET and its partners list drones, autonomous vehicles, robotic lawn mowers, construction equipment, agricultural machines, and industrial robots among the systems using this kind of correction data. (geodnet.com; quectel.com) GEODNET’s pitch is that it did not build this network the old way, with one company paying for every tower and receiver. It uses a decentralized model in which independent operators install “satellite miners” or reference stations on rooftops and feed data into GEODNET’s cloud platform in exchange for token rewards. (messari.io; geodnet.com) That model matters because RTK coverage gets better when stations are dense and standardized. Quectel, which partnered with GEODNET in May 2025, said GEODNET had more than 15,000 registered stations at the time and emphasized that every site used the same triple-band equipment, which makes it easier for manufacturers to test in one country and deploy in another. (quectel.com) The company’s growth has been fast by its own reported numbers. In February 2025, GEODNET said it had more than 13,500 user-deployed reference stations across 4,377 cities in 142 countries and that on-chain annual recurring revenue had grown more than 400% in 2024. (geodnet.com) By November 2025, Messari estimated GEODNET’s annualized revenue at about $5 million and said the network had expanded to more than 20,500 RTK base stations globally. As of this week, GEODNET’s own homepage shows figures above that, suggesting the business kept growing into 2026. (messari.io; geodnet.com) The customer story has also shifted toward robotics. GEODNET’s February 2025 funding announcement said the new capital would support robotics and “physical artificial intelligence” applications, and it specifically highlighted customers and partners such as Propeller, DroneDeploy, Quectel, the United States Department of Agriculture, Hemisphere GNSS, and Septentrio. (geodnet.com) That helps explain the road-versus-sidewalk example in the headline. A delivery drone, sidewalk robot, or humanoid machine does not just need to know the city block; it needs enough positional confidence to stay inside a narrow legal and physical corridor while cameras, lidar, and inertial sensors handle everything else around it. (geodnet.com) There is still a gap between the numbers in the social post and the numbers currently shown on GEODNET’s own site. The safest reading is that the post captured an earlier snapshot, while GEODNET’s live dashboard now shows a larger network and higher recurring revenue. (geodnet.com) The bigger point is simple: precision location has started to look less like a feature and more like infrastructure. If GEODNET can keep turning a crowd-built network of rooftop receivers into recurring software-like revenue from robots, it will have found a rare business in crypto-adjacent infrastructure that sells something factories, drones, and autonomous machines actually need in the real world. (messari.io; geodnet.com)