There's growing excitement around physical AI, fueled by Hansen Juang's keynote address at CES earlier this year and Andrew Kang's $19 million investment in Figure AI. There have been several focus areas within this field from a web3 perspective:
One of the key enablers of physical AI is spatial intelligence, which allows robots to understand, navigate, and interact with their environment. Over the Reality is among the leading players in this space, ranking as one of the world’s largest 3D mapping companies with over 112,000 maps—second only to Niantic, the creator of Pokémon Go, which boasts over 200,000 usable 3D maps. Adding to the intrigue, Niantic recently announced that it has sold its gaming division to focus solely on spatial intelligence.
This piece will provide an in-depth look at Over The Reality ($OVR) and why it could be an under the radar opportunity in the evolving spatial computing landscape, one of the critical building blocks of physical AI.
How OVER started
OVER was founded in 2018 by Davide Cuttini (CEO) and Diego Di Tommaso (COO). The company initially focused on AR (Augmented Reality) and digital land sales, drawing inspiration from Decentraland. The idea was to create an AR-based metaverse where users could buy virtual land in physical locations to publish content.
Over The Reality has garnered more than 1 million downloads across Android and iOS and has sold over 870,000 NFTs of digital land, generating more than $10 million in revenue. They previously launched a game similar to Pokémon GO, offering an end-to-end platform to publish AR content tied to physical locations. Initially, GPS was used to position content in space, but it quickly became clear that GPS lacked the precision required. The company pivoted to focus on a more promising solution: Virtual Positioning System (VPS).
Transition to VPS
The Virtual Positioning System (VPS) is a new paradigm for mapping and localizing within 3D environments. While GPS relies on satellite signals to provide global coordinates, VPS anchors devices based on visual landmarks captured from cameras and sensors within high-resolution 3D maps. This enables centimeter-level positioning and spatial awareness — essential for augmented reality, autonomous machines, and immersive digital experiences.
The technical breakthrough that led OVER to focus on VPS was 3D Gaussian Splatting, which gained attention in 2023 as a novel approach to real-time 3D scene reconstruction and rendering. The technique was introduced by researchers from Inria and Polytechnique in their paper "3D Gaussian Splatting for Real-Time Radiance Field Rendering". It quickly became a hot topic in the graphics and AI communities due to its ability to produce high-quality, photorealistic 3D reconstructions in real-time — something previous methods like NeRF (Neural Radiance Fields) struggled with due to computational cost.
OVER built its platform around this breakthrough, delivering precise AR experiences with VPS at its core. VPS relies on 3D maps, which are first required to map a location. In 2023, OVER pivoted from developing a full AR platform to building the infrastructure for mapping locations. Initially, OVER incentivized users through its Pokémon GO-like game, encouraging them to visit locations and earn tokens. However, the project later shifted those resources toward incentivising mappers.
What do they exactly mean by a map though? A map represents a 300-square-meter area, often referred to as a hexagon, based on the H3 library developed by Uber for geospatial calculations. To create a map, users physically visit a location and capture between 400 to 1,000 images with their smartphones. These images are then processed into a 3D structure, enabling VPS to position devices with up to 5 centimeters of precision.
Source: Over Marketplace
OVER Land and Incentivization
Each hexagon corresponds to OVER Land — a virtual parcel that grants publishing rights for AR content. If the location is mapped, AR experiences benefit from centimeter-level accuracy. If not, GPS-level accuracy applies. OVER plans to launch a revenue-sharing model where landowners will receive 35% of the revenues generated by services provided in their mapped hexagon, such as micro-mobility partnerships. Land prices range between $10 and $100, depending on the location's demand, with an average sale price of around $15.
Community
OVER has over 50,000 Monthly Active Users (MAU) and 12,000 new downloads monthly. Many contributors are digital nomads mapping locations as a side hustle, with top mappers generating up to 2,000 maps per month. Token rewards vary between $0.9 and $3 per map, depending on location importance and $OVR market price.
Supply Side Economics & Mapping Progress
OVER launched its mapping incentive program in September 2023, with map generation growing by 30% month-over-month. By late 2023, the program began prioritizing prime locations; they have since surpassed 112,000 maps.
Source: OVER Dune Dashboard
They are currently paying $1.25-1.30 per map generated and are now reaching 10,000 maps/month. The target is to reach 500,000 mapped locations to reach critical mass.
Use Cases
OVER is looking to cater five primary Web2 client segments, each benefiting from OVER’s Visual Positioning System (VPS) and geo-localized AR capabilities:
Digital Marketing Agencies & App Development Companies - OVER will offer a no-code platform that enables agencies to publish and browse white-label, geo-localized AR content effortlessly. This allows them to enhance their service offerings with immersive AR experiences, expanding their value proposition to brands and advertisers. The AR advertising market is valued at $5.8 billion in 2024 and projected to reach $8 billion by 2030. They have already done two proof of concept campaigns with Valentino and Pinko:
Additionally, OVER will provide an SDK that allows developers to seamlessly embed OVER’s AR and VPS capabilities into third-party applications. This integration improves precise re-localization and navigation within external apps. The global augmented reality navigation market size was estimated at USD 1.62 billion in 2024 and is expected to grow at a CAGR of 39.1% from 2025 to 2030. The no-code & SDK platform will mirror Niantic’s Spatial platform.
Micro-Mobility & Ride-Sharing Companies - Ride-sharing companies operate fleets of e-scooters and e-bikes under two models: dockless and docked. Docked systems rely on fixed urban parking stations, while dockless systems offer flexibility by allowing vehicles to be parked anywhere within designated zones. The dockless approach is a win for scalability but the disadvantage is its operational complexity.
Customers don’t always park in designated areas, and operators struggle to confirm proper parking using GPS alone. GPS accuracy hovers at 4-6 meters in open conditions and degrades to 10-15 meters in urban environments due to signal reflection off buildings. For a task requiring a maximum error of just 1 meter, GPS falls short.
Google employs VPS to enhance navigation in urban areas where GPS falters. The answer lies in precision. While Google’s VPS excels at broad navigation, it lacks the sub-meter accuracy ride-sharing companies need for parking verification. Google’s 3D maps, built from 360-degree imagery captured by its Street View cars every 10 meters or so, prioritizing breadth over depth. Whereas OVER takes a different approach takes 400-1,000 photos per 300 square meters, taken from multiple angles:
VIDEO: Google VPS vs OVER VPS: The Truth About Precision Mapping
This dense, diverse dataset delivers much better re-localization precision accurate to mere centimeters—relative to the designated parking zone.
What are the real cost savings of reducing this operation complexity?
Avoiding Fines: Improper parking can trigger steep penalties—think $150 per scooter in cities like San Francisco (SFMTA Fines).
Operational Savings: Fewer manual parking corrections mean lower labor expenses, saving thousands annually for large fleets.
Revenue Growth: Reliable parking improves customer satisfaction, boosting ride frequency and revenue potential.
Consider the following with a hypothetical ride-sharing company operating 1,000 scooters:
Assumptions:
5% of parking attempts are incorrect (125 daily errors at 2.5 rides/day/scooter).
$150 fine per violation.
$5 logistical cost per correction.
VPS reduces errors by 90%.
Avoiding fines: a company with 1000 ride-sharing vehicles will generate 2,500 rides per day, with a 5% incorrect parking rate we have 125 incorrect parkings daily. With a fine cost of $150 we have $18,750 in potential fines daily, $6,843,750 yearly. With a reduction of 90% in incorrect parking thanks to the VPS service there is a saving in potential fines for $6,159,375.
Of course not all of the infractions will be caught by local authorities, so let’s also assume that only 20% of incorrectly parked vehicles will actually get fined, the real savings in fines becomes $1,231,875.
Operational savings: All of the incorrectly parked vehicles, whether they get fined or not, will have to be physically moved to a correct parking location to become usable/reachable for the next client. Manual intervention for parking corrections is labor-intensive. Under the assumption of $5 logistic cost for each parking correction we have 125*365 = 45,625 parking corrections yearly for a total additional cost of $228,125 yearly. The 90% reduction in incorrect parking delivered by the VPS service would result in $205,312 in additional yearly savings of logistical costs.
The dockless ride-sharing market—spanning e-scooters and bikes—is booming. In 2025, it is projected to be valued at $2.7 billion, driven by urbanization and the growing demand for sustainable transport. By 2030, the market is expected to reach $6.4 billion, with e-scooter sharing growing at a CAGR of 18.8%. Europe leads the sector due to its e-scooter dominance, followed by North America and Asia-Pacific. Currently, parking verification using Google Maps VPS typically costs $0.05 per parking session. OVER aims to undercut Google with an aggressive pricing model of $0.02 per parking session to quickly capture market share. The average ride costs $3.5, translating to approximately 771 million rides annually. While $0.02 per ride may seem like a modest revenue stream, the operational cost savings and precision benefits offered by OVER’s VPS make it an attractive proposition for fleet operators. Once the market is captured, OVER could scale pricing, significantly increasing its revenue potential while still delivering strong value to customers.
Pilot projects are already underway with Forest which has a presence in London and Paris.
Industrial Facility & Real Estate Management - OVER’s AR-powered digital twins and VPS-driven indoor navigation offers businesses a powerful tool for maintenance, remote monitoring, and physical infrastructure management, enhancing operational efficiency.
Robotics Companies - Following the success of Large Language Models (LLMs), a new class of AI is emerging: Large Geo-Spatial Models (LGMs). LGMs will help computers perceive and comprehend the physical world in a way that will seem equally advanced. Like LLMs, geospatial models are built using vast amounts of raw data: billions of images of the world, all anchored to precise locations on the globe, are distilled into a large model that enables a location-based understanding of space, structures, and physical interactions. The shift from text-based models to those based on 3D data mirrors the broader trajectory of AI’s growth in recent years.
Unlike typical 3D generative models, which produce unscaled assets, a Large Geospatial Model is bound to metric space, ensuring precise estimates in scale-metric units. The combination of VPS and Large-Scale Geospatial Models (LGM) allows autonomous robots to precisely orient themselves in physical space and understand spatial composition via 3D segmentation, improving navigation and decision-making.
The Over team is currently working to develop a geospatial model with a university in Germany.
The OVER team is targeting $2M in revenue for 2025, with $1.5M expected from the no-code & SDK platform and $500K from micromobility API usage.
Competitors
Niantic is undoubtedly OVER's biggest direct competitor, but OVER holds several key advantages:
Consistent & Higher Quality: Most of Niantic’s data is collected through Pokémon GO, where the primary objective was never to construct a 3D map. As a result, while OVER may have fewer maps, its higher-quality datasets could lead to a superior Large Geospatial Model (LGM). In fact, its dataset is 1 to 2 orders of magnitude larger than those currently used by state-of-the-art Large Geospatial Models.
Incentivized Supply: OVER has incentivized data collection from day one, whereas Niantic relies on a mix of Pokémon GO users and its free-to-use app, Scaniverse. Incentivization typically results in better quality data and more targeted mapping efforts.
SDK Pricing: OVER plans to adopt a more aggressive pricing strategy, undercutting Niantic’s SDK rates by more than 50% to attract developers & marketing agencies.
Niantic’s last fundraise in 2021 valued the company at $9 billion, coinciding with the peak of its gaming division's revenue, which has been declining ever since. It has since sold its gaming division for $3.5 billion, and we estimate the company's total value to be around $4.5 billion. This suggests that Niantic Spatial could be valued at approximately $1 billion, especially considering the recent $250 million investment into the division.
In the web3 landscape, Meshmap is a notable competitor but follows a different approach. Meshmap relies on third-party scanning apps (e.g., Polycam, Scaniverse) rather than building its own mapping pipeline — a critical distinction, as this means it lacks access to raw image data, which is essential for delivering high-precision VPS.
Token
The OVER token is based on the ERC-20 standard, with its emission governed by a DAO. The current market cap is $6.1M, while the fully diluted valuation (FDV) stands at $10.4M. Most of the vested tokens remain in the foundation/company treasury, while the team’s allocation is mostly unlocked.
The token's utility revolves around supply incentivization and governance. Value capture will come from a portion of the revenue generated being used to buy back and burn tokens, with approximately 35% of the revenue distributed to OVER Land owners.
The current liquidity for the token is around $900,000 on Ethereum and the token is listed on exchanges like Kucoin and Gate.io.
Recent Fundraising Details
OVER recently raised $550,000 in an OTC round from well known DePIN investors such as Modular Capital, EV3 and Lattice.
Conclusion
Over The Reality ($OVR) presents a compelling bet in the spatial computing landscape at $10.4M FDV. With a growing network of high-quality 3D maps, incentivized data collection, and competitive pricing, Over is well-positioned to unlock use cases across digital marketing, micro-mobility, industrial management, and robotics — making it a standout player at the intersection of physical AI and web3.
