{"version":1,"id":"edgeai","title":"EdgeAI","summary":"EdgeAI is a Layer1 project that builds a blockchain-enabled network to enable AI computation and data monetization at the edge, combining federated learning, differential privacy and a Proof of Information Entropy (PoIE) consensus mechanism.","content":"**EdgeAI** is a [blockchain](https://iq.wiki/wiki/blockchain) project that describes itself as a [Layer 1](https://iq.wiki/wiki/layer-1) network for artificial intelligence at the network edge, combining decentralized physical infrastructure with on-device computation. The project presents itself as \"The Most Intelligent Data Chain for Edge AI,\" a secure, scalable, and efficient Layer 1 solution that [leverages](https://iq.wiki/wiki/leverage) [DePIN](https://iq.wiki/wiki/depin) — decentralized physical infrastructure networks, a model in which individually owned hardware is coordinated by a blockchain — to run AI processing on devices near where data is produced rather than in centralized clouds.[[1]](#cite-id-lfoocxk0h1) Its self-described purpose is to keep inference and decision-making close to the data source so that latency and privacy exposure are reduced.[[2]](#cite-id-61an89lxky) The project's X account, @EdgeAI_xyz, was created in May 2024, and it publishes an [Ethereum](https://iq.wiki/wiki/ethereum) address, 0x276b792d11b9e3712fe6a78a460a0deb416bab0a, on its profile.[[2]](#cite-id-61an89lxky)\n\n## Overview\n\nEdgeAI frames the problem it addresses as the growing volume of data produced by connected devices, which it says exceeds the capabilities of traditional systems and leads to delays and privacy risks.[[1]](#cite-id-lfoocxk0h1) Its stated approach is to merge edge computing, differential privacy, and federated learning within a [blockchain](https://iq.wiki/wiki/blockchain) framework, a combination the project claims enables real-time decision-making, stronger data privacy, and secure data monetization.[[1]](#cite-id-lfoocxk0h1) In its own words, \"Most AI still ships data to the cloud, waits, then guesses. EdgeAI flips that: inference happens where the data is created; privacy stays local; valuable contributions get measured, not just counted; devices can settle value as they work.\"[[2]](#cite-id-61an89lxky)\n\nThe project organizes its focus areas into four categories: edge computing integration, data sovereignty and privacy, high-performance blockchain, and data monetization opportunities.[[1]](#cite-id-lfoocxk0h1) It also lists performance goals including energy-efficient data processing, reduced latency, and efficient AI learning at the edge.[[1]](#cite-id-lfoocxk0h1) EdgeAI publishes a [whitepaper](https://iq.wiki/wiki/white-paper) and distributes a pitch document via DocSend under the EdgeAI title; contact for the project is directed to support@edgeai.xyz.[[3]](#cite-id-mcvbc3b4ot)[[1]](#cite-id-lfoocxk0h1)\n\n## Technology and Architecture\n\nEdgeAI's consensus design centers on a mechanism it calls **Proof of Information Entropy (PoIE)**. The project states that PoIE evaluates and rewards high-quality data contributions from Internet-of-Things (IoT) devices, structuring the network's [mining](https://iq.wiki/wiki/mining) so that participating hardware is compensated for the informational value it supplies rather than for raw computational effort.[[1]](#cite-id-lfoocxk0h1) The project has said PoIE is built to let devices \"settle value as they work\" and to have contributions \"measured, not just counted.\"[[2]](#cite-id-61an89lxky)\n\nThe project describes a privacy-first architecture built from three named components. Local differential privacy adds statistical noise to data on the device itself before it is shared; federated learning with differential privacy trains shared models across many devices without moving the underlying raw data to a central server; and privacy-preserving data aggregation combines contributions in a form intended to protect individual records.[[1]](#cite-id-lfoocxk0h1) EdgeAI positions these techniques as the means by which it can process data at the edge while keeping sensitive information local, supporting what it describes as secure data monetization.[[1]](#cite-id-lfoocxk0h1) All of these architectural elements are as described by the project.\n\n## EdgeML and TinyML Computer Vision\n\nEdgeAI maintains a public code repository, 0xEdgeAI/EdgeML, created on 28 October 2025, whose top-level project is titled \"TinyML-Computer-Vision.\"[[4]](#cite-id-w1r251sfhk) The repository presents a smart-doorbell use case built around edge-based image recognition, listing target detections that include faces, pets such as cats and dogs, unsafe content such as guns, noteworthy vehicles such as ambulances and firetrucks classified as trucks, cars, courier company logos (USPS, FedEx, DHL, and Amazon Prime), and delivery packages or courier boxes.[[4]](#cite-id-w1r251sfhk)\n\nThe repository documents three distinct technical approaches to these detection tasks. The first is a cloud software-as-a-service (SaaS) pipeline built on Amazon Web Services (AWS), which the project describes as implementing scene-detection analytics across edge and cloud resources. This approach uses AWS Rekognition for scene detection in video, AWS Kinesis to upload analytics data from the edge to the cloud, AWS DynamoDB to store that data, AWS S3 to store videos and image frames, and AWS SNS for service notifications.[[4]](#cite-id-w1r251sfhk) The code for this approach is organized around a main program that draws on modules for AWS credentials, Rekognition functions, and video capture, along with a script for indexing known faces so that recognized visitors can be identified at a doorstep.[[4]](#cite-id-w1r251sfhk) The project states that this approach attempts to reproduce the scenario described in the paper \"Demonstration of a Cloud-based Software Framework for Video Analytics Application using Low-Cost IoT Devices.\"[[4]](#cite-id-w1r251sfhk)\n\nThe second approach runs computer vision directly on-device, targeting hardware including the Raspberry Pi 3 Model B+, the NVIDIA Jetson [Nano](https://iq.wiki/wiki/nano), and the Arduino Nano 33 BLE. It relies on the TensorFlow Lite interpreter to run several models: a weapon-detection model (gun_model_2.2.tflite), a courier-logo detection model for DHL, FedEx, Amazon Prime, and USPS logos on delivery vans (LogoModel.tflite), a pre-trained MobileNet SSD v2 model for common objects such as people, pets, cars, and noteworthy vehicles (MobileNetV2.tflite), and a package-detection model for parcels delivered to doorsteps (PackageModel.tflite).[[4]](#cite-id-w1r251sfhk) The repository organizes datasets, TFLite models both pre-trained and custom-trained, label maps defining class names, and result folders containing detection output from experiments run on the Raspberry Pi and Jetson Nano.[[4]](#cite-id-w1r251sfhk)\n\nThe third approach uses classical computer vision through the Viola–Jones algorithm and Haar cascade classifiers, an older technique that detects objects using patterns of light and dark regions. This approach ships a set of trained cascade files for the same doorbell use cases — separate XML classifiers for frontal faces, guns, cats, dogs, ambulances, firetrucks, packages, and the Amazon, DHL, FedEx, and USPS logos.[[4]](#cite-id-w1r251sfhk) The repository also references sections for comparative analysis, an API server, and a mobile app.[[4]](#cite-id-w1r251sfhk)\n\nThe repository ties its work to academic dissemination, listing three research papers: \"A Distributed Framework to Orchestrate Video Analytics Applications,\" \"Demonstration of a Cloud-based Software Framework for Video Analytics Application using Low-Cost IoT Devices,\" and \"A Demonstration of Smart Doorbell Design Using Federated Deep Learning.\"[[4]](#cite-id-w1r251sfhk) It also references a demonstration video titled \"A Cloud-based Smart Doorbell using Low-Cost COTS Devices.\"[[4]](#cite-id-w1r251sfhk)\n\n## Partnerships and Community\n\nEdgeAI states that its network spans more than 10,000 edge devices, a figure it uses to describe the scale at which its data monetization and inference services would operate.[[2]](#cite-id-61an89lxky) In August 2026, the project announced a partnership with Solulu Pay (@SoluluPay) to bring what both parties described as real-time, privacy-first crypto micropayments to that device network, with the stated aim of enabling data monetization and AI inference \"at machine speed.\"[[2]](#cite-id-61an89lxky) A corresponding Solulu Pay post described the collaboration as connecting [Web3](https://iq.wiki/wiki/web3) payment rails with EdgeAI's device network to enable real-time crypto settlement for edge data monetization and AI inference with a privacy-first checkout.[[2]](#cite-id-61an89lxky)\n\nThe project has also participated in several community events. On 18 August 2026 it joined a Tilted (@tiltedxyz) community AMA in Discord under the theme \"Where AI Platforms Meet the Edge,\" and around the same period it took part in a WikiBit (@WikiBitOfficial) X Space scheduled for 20 August 2026, which was promoted with a $100 USDT giveaway.[[2]](#cite-id-61an89lxky) EdgeAI additionally announced joining a Space with Beatcoin (@BrcToTheMoon) to discuss the [Bitcoin](https://iq.wiki/wiki/bitcoin) rally, ETF flows, and institutional positioning.[[2]](#cite-id-61an89lxky)\n\n## Recognition\n\nEdgeAI states that it was awarded the Technology [Frontier](https://iq.wiki/wiki/frontier) Award by the Bincheng International Innovation Competition.[[1]](#cite-id-lfoocxk0h1) This distinction is reported by the project itself.","categories":[{"id":"projects-and-protocols","title":"projects-and-protocols"}],"tags":[{"id":"Layer1"},{"id":"Infrastructure"},{"id":"AIPlatform"},{"id":"AIInfrastructure"},{"id":"Ethereum"}],"images":[{"id":"QmUXget8As112TF85dmBV4Z9knNUbnKeSq6Ma9zGV6rfAn","type":"image/jpeg, image/png"}],"media":[],"metadata":[{"id":"references","value":"[{\"id\":\"lfoocxk0h1\",\"url\":\"https://edgeai.xyz/\",\"description\":\"EdgeAI Layer 1 and DePIN self-description\",\"timestamp\":1788380776076},{\"id\":\"61an89lxky\",\"url\":\"https://x.com/EdgeAI_xyz\",\"description\":\"EdgeAI distributed intelligence network bio\",\"timestamp\":1788380776076},{\"id\":\"mcvbc3b4ot\",\"url\":\"https://docsend.com/view/eber7qywukyavpxm\",\"description\":\"EdgeAI DocSend document and support email\",\"timestamp\":1788380776076},{\"id\":\"w1r251sfhk\",\"url\":\"https://github.com/0xEdgeAI/EdgeML\",\"description\":\"EdgeML repository creation date and title\",\"timestamp\":1788380776076}]"},{"id":"rootdata_profile","value":"https://www.rootdata.com/Projects/detail/EdgeAI?k=MjI2NzM="},{"id":"website","value":"https://edgeai.xyz/"},{"id":"twitter_profile","value":"https://x.com/EdgeAI_xyz"},{"id":"medium_profile","value":"https://medium.com/@EdgeAI2024"},{"id":"coinmarketcap_url","value":"https://coinmarketcap.com/currencies/39004/"},{"id":"youtube_profile","value":"https://www.youtube.com/watch?v=42mx4Z2PDwA"},{"id":"telegram_profile","value":"https://t.me/EdgeAI_blockchain"},{"id":"commit-message","value":"Create EdgeAI wiki"}],"events":[{"type":"DEFAULT","title":"EdgeAI X account joined","date":"2024-05-01","description":"EdgeAI account on X (Twitter) joined.","link":"https://x.com/EdgeAI_xyz","multiDateStart":null,"multiDateEnd":null,"id":"4d61e6e9-6e6f-4c8b-b5f9-9c7aa8a32064"},{"type":"DEFAULT","title":"0xEdgeAI/EdgeML repository created","date":"2025-10-01","description":"Repository 0xEdgeAI/EdgeML created on GitHub.","link":"https://github.com/0xEdgeAI/EdgeML","multiDateStart":null,"multiDateEnd":null,"id":"4087b799-3d5d-4999-866d-cf33efff9d3f"},{"type":"DEFAULT","title":"WikiBit X Space participation announced","date":"2026-08-01","description":"Announcement of participation in a WikiBit X Space.","link":null,"multiDateStart":null,"multiDateEnd":null,"id":"4613ead6-3e17-40f9-9208-85e4a092cd40"}],"linkedWikis":{"founders":[],"blockchains":["ethereum"],"speakers":[]},"user":{"id":"0x8af7a19a26d8fbc48defb35aefb15ec8c407f889"},"author":{"id":"0x2D403Be9b67bE06FFEEb7C2681e19FB4D0eeA48b"},"language":"en","operator":{"id":"0x2D403Be9b67bE06FFEEb7C2681e19FB4D0eeA48b"}}