Portfolio / AIoT Edge Inspection Station
2025 Edge AI Computer Vision LiDAR Raspberry Pi

AIoT Edge
Inspection Station

A real-time visual inspection system on Raspberry Pi 5 combining YOLOv6n object detection at ~28 FPS on a Hailo-8L NPU with LiDAR-based proximity sensing — distance overlaid on every bounding box in real time.

Hardware-accelerated AI meets multi-sensor fusion.

Developed as part of an MSc course in AIoT Edge Computing at FH Technikum Wien, this system demonstrates how commodity edge hardware — a Raspberry Pi 5 with a Hailo-8L neural processing unit — can run production-grade object detection at real-time speeds while simultaneously fusing LiDAR range data to provide spatial context on every detection.

The system evolved across three milestones: starting with bare NPU detection and LiDAR fusion, adding a proximity-triggered snapshot system with stateful cooldown, and finishing with face recognition, a Flask web dashboard, and Telegram security alerts for unknown visitors.

Three cooperating processes

Camera (CSI)
RPi Camera v2/v3
Hailo-8L NPU
YOLOv6n ~28 FPS
Sensor Fusion
sensor_fusion.py
LiDAR Server
UDP JSON broadcast
Hokuyo LiDAR
URG-04LX USB serial

MILESTONE 1 — DETECTION

YOLOv6n at ~28 FPS on Hailo-8L. Distance from LiDAR overlaid on every bounding box. Real-time LiDAR polar map rendered alongside.

MILESTONE 2 — PROXIMITY GATE

LiDAR monitors entry zones and triggers camera snapshots with stateful cooldown to prevent re-triggering events.

MILESTONE 3 — FULL SYSTEM

Face recognition (dlib) classifies known vs. unknown visitors. Flask dashboard for live monitoring. Telegram alerts on detections.

EVENT LOGGING

All detections logged to rotating CSV files with timestamps and confidence scores for post-analysis.

Components

COMPONENTSPECIFICATION
ComputeRaspberry Pi 5 (4–8 GB RAM)
AI AcceleratorHailo-8L M.2 neural processing unit
LiDARHokuyo URG-04LX-UG01 (USB serial)
CameraRaspberry Pi Camera Module v2/v3 (CSI)
ModelYOLOv6n — ~28 FPS on Hailo-8L
OSRaspberry Pi OS 64-bit Bookworm
Proximity thresholdConfigurable — default 1.5 m

Technologies used

LANGUAGES
Python Shell
AI / VISION
Hailo-8L YOLOv6n OpenCV dlib face_recognition
BACKEND
Flask NumPy
ALERTS
Telegram Bot API CSV Logging