XYO integrates DePIN network with NVIDIA Jetson Orin Nano

XYO's blockchain-signed sensor pipeline now runs natively on NVIDIA's Jetson Orin Nano, giving edge devices a tamper-evident record of what they

A ruggedized embedded computer system with exposed circuit boards and multiple colored cables is on a white desk in a bright, modern office or laboratory.

XYO, the San Diego-based decentralised physical infrastructure network (DePIN) operator, has built a full end-to-end integration that runs on NVIDIA's Jetson Orin Nano system-on-module. The integration allows any device carrying the chip to cryptographically sign sensor observations at the point of capture and write them as permanent records to XYO Layer One, the company's purpose-built blockchain. XYO reported $8.8 million in 2024 revenue and claims more than ten million active nodes across its network.

The technical pipeline works entirely on NVIDIA hardware. A camera or other sensor feeds raw input to the Jetson Orin Nano, which runs computer vision inference locally, classifies the observation, attaches a confidence score, and signs the event before relaying it to XYO Layer One. No cloud hop is required between capture and the on-chain record, which the company says prevents third parties from modifying or deleting the data in transit.

The case for verifiable sensor data

XYO frames the integration as a direct response to a structural gap in current autonomous systems: sensor data that sits on private servers, is invisible to other agents in the same fleet, and can be altered or erased before any post-incident audit. Co-founder Markus Levin pointed to the accountability problem directly: "Autonomous systems are already making decisions on sensor data they have no way to verify. It sits on private servers where it can be edited or deleted, and the machine acting on it has to take it on faith. This integration signs what a device detects at the point of capture and writes it to a record that cannot be rewritten, so the data can be traced back to its origin."

The company illustrated the pipeline with a warehouse and agriculture framing: a thermal imager signing maintenance-log entries when a motor runs hot, a depth camera proving a pallet left a loading dock at a specific time, or a drone producing crop-condition data a lender or insurer can act on remotely. XYO also described a domestic demonstration in which a camera attached to a Jetson Orin Nano classified a dog entering a room with 86.4 per cent confidence and triggered a treat dispenser, with the complete chain of events recorded on-chain and no keyboard input required.

NVIDIA's Jetson Orin Nano ecosystem reportedly spans more than 150 hardware partners, 2,000 startups and a large base of commercial customers across warehousing, logistics and agriculture. XYO's next announced milestone is support for NVIDIA Isaac ROS, the GPU-accelerated implementation of the ROS 2 robotics standard that runs natively on Jetson hardware.

Market context and competitive positioning

The DePIN category has grown rapidly since 2022, attracting projects that seek to monetise real-world infrastructure by tokenising data contributions. XYO positions itself as the original DePIN, having operated since 2018, but the space now includes a number of well-funded competitors building location, compute, storage and sensor sub-networks. The specific combination of edge inference, cryptographic signing and on-chain storage is less crowded, but it touches multiple established markets: IoT data platforms (where AWS IoT, Azure IoT Hub and Google Cloud IoT compete), blockchain data-availability layers, and robotics middleware.

From a regulatory standpoint, the audit-trail argument XYO makes is relevant to emerging frameworks for autonomous systems liability. The EU's AI Act and the proposed EU Product Liability Directive both place increased evidentiary obligations on operators of high-risk autonomous systems, which could create commercial demand for tamper-evident sensor logs of exactly the kind XYO describes. In the UK, the Automated Vehicles Act 2024 similarly establishes a framework that will require incident-data records for self-driving vehicles. Whether XYO's on-chain approach satisfies the evidentiary standards these regimes will impose remains to be tested, and the company has not cited formal certification against any such framework.

Investors and enterprise buyers will be watching for named commercial deployments at scale, independent benchmark data on throughput and storage costs at high node density, and the Isaac ROS release as the near-term proof points for this integration.