soaring-xiongkulu/easyaiot

EasyAIoT – An All‑In‑One AI‑Powered IoT Platform

What it is – EasyAIoT (Easy AI Internet of Things) is a single‑stack software platform that ties together cloud services, edge compute boxes and end‑devices (cameras, sensors, industrial PLCs, etc.) so they can share data and run AI models together. It is built to run on very small edge boxes (≈1 GB RAM) up to full‑size enterprise appliances.

Core capabilities

• Unified device onboarding – supports GB28181, ONVIF and a wide range of consumer cameras (Tapo, Tuya, Ring, Nest, Xiaomi, Wyze, DoorBird, GoPro, Roborock) via an RTC bridge built on go2rtc that turns proprietary P2P streams into standard RTSP.
• Real‑time video AI – native runtime binary performs YOLO object detection, SAM zero‑shot annotation, face‑plate recognition and other vision tasks with low CPU/memory overhead.
• Post‑processing rules – after detection you can filter, combine, or transform results into business events without retraining the model.
• Edge‑to‑cloud orchestration – a “Sentinel” scheduler watches node health, distributes inference jobs, and can grow into an infinite federated edge cluster (development boards become compute nodes automatically).
• Industrial protocol support – built‑in C# collectors for Modbus‑TCP/RTU, OPC UA and MQTT, exposing device telemetry to the same console.
• Visualization & SCADA – dashboards, map‑based situational awareness and configurable data‑flow engine (TRANSFORM) that pushes events to MES/ERP/CRM/WMS.
• Cross‑platform packaging – installers for Ubuntu/Debian, CentOS/RHEL 7‑9 (x86 & ARM), Kylin, openEuler, Windows, macOS and HarmonyOS; Docker containers are also provided.
• Mobile & web UI – a responsive web console plus Android/iOS/HarmonyOS apps built from a single front‑end codebase; a “PANEL” tool lets field engineers install, watch health, and troubleshoot appliances on‑site without remote developer help.

Typical use cases

• Small‑scale smart‑store video monitoring that re‑uses cheap consumer cameras and adds AI alerts.
• Industrial edge gateways that collect sensor data, run defect‑detection models locally, and forward only events to the cloud.
• Drone or DJI‑dock video feeds that are analysed in real time for security or inspection.
• Multi‑site surveillance where each site runs a lightweight edge node and the cloud coordinates model updates and analytics dashboards.

Getting started (Linux)

# clone the repo
git clone https://gitee.com/volara/easyaiot.git
cd easyaiot

# install the edge profile (Docker‑based)
EASYAIOT_DEPLOY_PROFILE=edge sudo bash .scripts/docker/install_linux.sh install

# verify and open the console
.scripts/docker/install_linux.sh verify
# then browse to https://<your‑server‑ip>:8888 (admin / admin123)

The installer reports memory usage (≈1 GB for the lightest tier) and checks that all containers are healthy.

Why it matters – Most IoT deployments struggle because video, device management and AI are handled by separate products, leading to costly integration work and fragile pipelines. EasyAIoT collapses that stack into one codebase, letting developers and integrators focus on the business logic rather than plumbing. Its emphasis on tiny edge footprints and plug‑and‑play camera bridging makes AI‑enhanced vision affordable for small shops, factories, or hobby projects.

Where to learn more – The project ships an illustrated PPT, a PDF overview, and a full operations manual in the ./.doc/ folder, plus a public website (http://36.111.47.113:8090/) with installer downloads and documentation.