v1.0.0 100% Free · Open Source · MIT License

The brain for
physical spaces

Simulate IoT devices, process real-time data streams, run AI inference on CPU, make autonomous decisions, and control actuators — all locally, all free, all open-source.

# Clone & run in under 2 minutes
git clone https://github.com/rudra496/EdgeBrain.git
cd EdgeBrain
docker compose up --build -d
# ✅ Open http://localhost:3000 — done.
11
IoT Devices
3
AI Agents
20+
API Endpoints
37
Tests Passing
$0
Cost to Run

Everything you need. Nothing you don't.

A complete edge intelligence platform — from simulated sensors to autonomous decisions — with zero external dependencies.

📡

Smart Device Simulation

11 virtual IoT devices across 3 rooms with realistic circadian patterns, Gaussian noise, and rare event simulation. Temperature, motion, energy, humidity, and light.

🤖

Multi-Agent AI System

Data Agent validates and stores. Decision Agent evaluates strategies. Action Agent executes commands. Full message bus with performance tracking.

🧠

Anomaly Detection

Three statistical methods — Z-Score, IQR, and Gradient — vote on anomalies. Requires 2+ methods to agree, reducing false positives by ~70%.

📈

Prediction Engine

Linear regression, Simple Moving Average, and Exponential Moving Average forecasting. Multi-step predictions with confidence scores. CPU only.

📊

Real-Time Dashboard

React 18 with WebSocket-powered live charts, 5 navigable pages, agent pipeline visualization, and a polished dark theme design system.

🔌

REST + WebSocket API

20+ endpoints with Pydantic validation, auto-generated Swagger docs, CSV/JSON data export, device predictions, and bidirectional WebSocket.

📏

Rule Engine + Hysteresis

Configurable threshold triggers with hysteresis to prevent actuator flapping. No-motion timeout automatically turns off unused lights.

🧩

Plugin Architecture

Add custom decision strategies with 10 lines of Python. Just implement evaluate() and name — the engine handles the rest.

ESP32 Hardware Support

C++ Arduino firmware for real DHT11 temperature and PIR motion sensors. LED and buzzer actuator control. WiFi + MQTT.

From sensor to action, locally.

An event-driven pipeline that never leaves your machine. Every component is modular, testable, and replaceable.

1

IoT Sensors

11 devices, 3 rooms, MQTT stream

2

MQTT Broker

Mosquitto topic routing

3

Data Agent

Validate + store in PostgreSQL

4

Decision Agent

Rules + anomaly + prediction

5

Action Agent

Alerts + MQTT commands

6

Actuators

Lights, fans, alarms

7

Dashboard

Live WebSocket charts

See it in action

Three complete demos run automatically when you start the system. Watch AI make decisions in real-time.

Smart Room Automation

Intelligent room control that adapts to occupancy and environment conditions without any human input.

Motion → Lights ON5 min idle → Lights OFFTemp > 30°C → Fan ONTemp < 25°C → Fan OFF

Anomaly Detection

Statistical monitoring that catches equipment anomalies before they become critical failures.

Z-Score AnalysisIQR MethodGradient DetectionMulti-vote Consensus

Energy Monitoring

Continuous consumption tracking with time-of-day awareness, spike detection, and trend forecasting.

Usage PatternsSpike DetectionTrend ForecastingExport CSV/JSON

Built with the best tools

Every dependency is free, open-source, and battle-tested in production.

Python 3.11+
FastAPI
PostgreSQL 16
Redis 7
Mosquitto MQTT
React 18
Recharts
NumPy + SciPy
Docker Compose
Pytest
ESP32 / Arduino
GitHub Actions CI

Frequently Asked Questions

Everything you need to know about EdgeBrain.

What is EdgeBrain?
EdgeBrain is a free, open-source AI-powered edge intelligence platform that simulates IoT devices, processes real-time data, runs AI inference on CPU, makes autonomous decisions, and controls actuators — all running locally on your machine with zero cloud dependencies.
Is EdgeBrain completely free?
Yes, 100%. EdgeBrain uses only open-source tools (Python, FastAPI, PostgreSQL, Redis, Mosquitto, React) and requires no paid APIs, no cloud services, and no proprietary dependencies. It runs entirely on your local machine using Docker.
How do I set up EdgeBrain?
Install Docker and Docker Compose, then run three commands: git clone https://github.com/rudra496/EdgeBrain.git, cd EdgeBrain, docker compose up --build -d. Open http://localhost:3000 — that's it. Setup takes under 2 minutes.
What AI features does EdgeBrain include?
EdgeBrain includes a multi-agent AI system (Data Agent, Decision Agent, Action Agent), rule-based threshold triggers with hysteresis, statistical anomaly detection using 3 methods (Z-Score, IQR, Gradient), linear regression prediction engine, and a plugin architecture for custom strategies.
Can I use real hardware with EdgeBrain?
Yes! EdgeBrain includes ESP32 firmware (C++/Arduino) for real DHT11 temperature and PIR motion sensors. The firmware communicates via MQTT and supports LED and buzzer actuators. You can also connect any MQTT-compatible device.
How does the anomaly detection work?
Three independent statistical methods analyze sensor data: Z-Score (deviation from mean), IQR (interquartile range outliers), and Gradient (sudden changes between readings). Two or more methods must agree to trigger an alert, reducing false positives by approximately 70%.
Can I add my own custom logic?
Absolutely. EdgeBrain has a plugin architecture — just create a Python class that implements evaluate() and name methods, then register it with the decision engine. Custom strategies are evaluated alongside built-in ones and can trigger alerts and actuator commands.

No cloud required

Unlike IoT cloud platforms, EdgeBrain runs entirely on your machine. Your data never leaves your laptop.

FeatureEdgeBrainTypical IoT Cloud
Cost$0 forever$10–500+/mo
Data Privacy100% localCloud-stored
Setup Time~2 minutesHours to days
AI InferenceBuilt-in (CPU)Varies / extra cost
Custom StrategiesPython plugin systemLimited / locked
Hardware SupportESP32 includedVendor-specific
Internet RequiredNoAlways
Open SourceMIT LicenseProprietary
Latency< 5ms (local)100–500ms (cloud)

Built by Rudra Sarker

A Robotics and IoT developer specializing in AI, embedded systems, and social impact innovation.

Rudra Sarker

Robotics & IoT Developer · Engineering Student at SUST

Specializing in AI, assistive technology, and innovative engineering solutions. Building open-source tools that make edge intelligence accessible to everyone.

Ready to put a brain
on your edge?

One command. Zero cost. Full control.