Build a Retail Footfall Camera That Runs AI at 13 TOPS — Raspberry Pi 5 + Hailo-8L 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.

Imagine a camera that not only watches a store entrance but instantly classifies every person crossing the threshold — without sending a single frame to the cloud. This kit lets you build exactly that. You will mount a Hailo-8L AI accelerator onto a Raspberry Pi 5, connect the Camera Module 3, and deploy a MobileNetV2 image classification model to track footfall in real time. The real challenge is not just assembling hardware; it is running an inference benchmark and comparing the blazing speed of the Hailo-8L at 13 TOPS against the Pi 5’s CPU-only performance, a hands-on lesson in edge computing that transforms how you think about AI at the point of capture.

What You'll Build

You will create a self-contained edge AI camera that captures live video, classifies each frame as containing a person or not, and logs timestamped counts to the onboard NVMe SSD. The system runs entirely on-device, making it a miniaturised version of the people-counting analytics used in modern retail. By the end of the build, you will have a working footfall monitor that you can benchmark, tweak, and even extend to recognise other objects.

What You'll Learn

• Deploy a pre-trained MobileNetV2 model on the Hailo-8L neural processor and measure its throughput in frames per second
• Compare inference performance between the Hailo accelerator and the Raspberry Pi 5’s Arm CPU using identical input streams
• Stream and process camera data with OpenCV and the HailoRT runtime on Raspberry Pi OS
• Design a real-world edge AI pipeline that captures, classifies, logs, and retrieves footfall metrics without cloud dependency

Kit Contents

Why Buy This Kit Instead of Sourcing Parts Separately

Who This Kit Is For

If you are a B.Tech ECE or CSE student exploring embedded vision for a Smart India Hackathon project, or a CBSE Class 12 student preparing a computer science investigatory project, this kit gives you a complete edge AI prototyping platform. It also suits ATL Tinkering Lab mentors who want to demonstrate real-time neural network inference on low-power devices. The build complexity aligns well with first-year engineering workshops at IITs, NITs, and VIT.

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.