AI-Assisted Presence and Behaviour Detector Kit with ESP32
Build a Room Occupancy Classifier with On-Device AI – ESP32 TensorFlow Lite Kit
Every part needed, pre-tested for compatibility, with an AI build companion trained on this exact project. Shipped from Bengaluru in 3-5 days.
Transform a simple ESP32-CAM into a silent, always‑on AI brain that knows exactly how a room is being used—without ever sending data to the cloud. By combining camera snapshots, passive infrared motion, and door usage patterns, you’ll train a compact TensorFlow Lite model that classifies occupancy in real time: vacant, single occupant, multiple occupants, or transition. Perfect for energy‑saving home automation, elder‑care monitoring, or a privacy‑respecting smart security layer.
What You'll Build
You’ll assemble and program a wall‑mountable sensor module that fuses three sensor streams into a single occupancy inference. Every few seconds the ESP32‑CAM captures a low‑resolution image, reads two PIR sensors for motion direction, and monitors reed switches on doors. The on‑device TensorFlow Lite model processes the combined feature vector and outputs a high‑confidence occupancy state. That state can trigger lights, HVAC, or alerts through MQTT—fully local, no internet required.
What You'll Learn
- Fusing multi‑sensor data (PIR, magnetic reed switch, ESP32‑CAM) into a unified feature vector for an AI model
- Capturing and preprocessing images from the ESP32‑CAM for on‑device inference at multiple frames per second
- Training a compact TensorFlow Lite model for 4‑class occupancy classification and converting it to run within the ESP32’s memory limits
- Deploying the model to the ESP32 and accelerating inference using hardware‑specific operators and careful buffer management
Kit Contents
| Component | Quantity |
|---|---|
| ESP32-CAM | 1 |
| HC-SR501 PIR | 2 |
| Reed Switch | 3 |
| Active Buzzer | 1 |
| LM2596 Buck Converter | 1 |
| LM1117 3.3V Reg | 2 |
| 1000µF 25V Caps | 2 |
| 100nF Caps | 10 |
| 10kΩ Resistors | 5 |
| PCB Prototype Board | 2 |
| Enclosure Box | 1 |
| 5V 2A Power Supply | 1 |
| Soldering Iron | 1 |
| Solder Wire | 1 |
Why Buy This Kit Instead of Sourcing Parts Separately
| Factor | Sourcing Separately | Compoden Kit |
|---|---|---|
| Compatibility checks | You verify every part | Pre-tested as a system |
| Build support | Forums and scattered tutorials | AI companion trained on this exact project |
| Time to first working build | Days of debugging | Hours, with step-by-step guidance |
| Shipping coordination | Multiple sellers, multiple delays | One shipment from Bengaluru in 3-5 days |
Who This Kit Is For
Engineers and data scientists exploring TinyML for real‑world environments, B.Tech ECE/CS final‑year students prototyping AI‑powered assistive tech for Smart India Hackathon, and experienced hobbyists ready to move beyond Arduino into embedded machine learning. If you’ve wanted to run a neural network on a microcontroller without a cloud dependency, this kit gives you a polished, end‑to‑end project that plugs straight into Home Assistant or any MQTT‑based automation.
Built and Backed by Compoden
Every Compoden kit ships with an AI build companion trained on this exact project — accessible via a QR code on the box, with WhatsApp and email backup. We've spent 10 years building projects for makers, schools, and institutions across India. If a part fails because of a manufacturing defect, replace it free within 7 days.
What if I get stuck during the build?
Scan the QR code inside the box to chat with the AI companion, which knows every wiring connection and code snippet for this kit. If you need a human, WhatsApp our support team—we typically respond within a few hours.
Can I modify the model to detect other states like sleeping or cooking?
Yes. The companion walks you through collecting new labeled data in your room and retraining the TensorFlow Lite model, so you can define any custom occupancy categories you need.
Will this work with Home Assistant or Node‑RED?
Absolutely. The ESP32 publishes the occupancy state via MQTT, which natively integrates with Home Assistant, Node‑RED, and other platforms for fully local automation.
How accurate is the occupancy detection in a real home?
After you calibrate the sensors to your room layout, the model typically exceeds 92% accuracy in distinguishing vacant, single‑person, and crowd conditions. It runs entirely on the ESP32, so there’s no internet latency or privacy concern.
ESP32-CAM + PIR + door sensors feed an on-device TensorFlow Lite model that classifies room occupancy state.
What's in this kit
- ESP32-CAM
- HC-SR501 PIR x2
- Reed Switch x3
- Active Buzzer
- LM2596 Buck Converter
- LM1117 3.3V Reg x2
- 1000µF 25V Caps x2
- 100nF Caps x10
- 10kΩ Resistors x5
- PCB Prototype Board x2
- Enclosure Box
- 5V 2A Power Supply
- Soldering Iron
- Solder Wire
Choose your assembly option:
- Soldering Kit — 25W soldering iron, 60/40 solder wire, flux, and small perfboard for permanent assembly.
- Breadboard Combo — 800-point full-size breadboard with 65-piece jumper wire pack for solderless prototyping.
Other projects you can build
Shipping Information
- Prepaid Orders: ₹75 for orders up to ₹999, FREE shipping above ₹999
- COD Orders: ₹125 shipping + ₹50 COD fee = ₹175 total
- Delivery Timeline: Dispatch in 1-2 days, delivery in 2-7 days depending on location
Returns & Warranty
- 7-Day Return: Manufacturing defects only (approval required)
- Warranty: 7 days from delivery
- Non-Returnable: Batteries, consumables, cut wires, clearance items