◆ AI QUALITY INSPECTION
UNITS SCANNED / HR
2,400
DEFECT DETECTION RATE
97.4%
FALSE REJECT RATE
1.2%

Project Overview

Deployed on a small manufacturing line, this system uses an industrial camera and a trained defect-detection model to inspect units as they pass on the conveyor, automatically triggering a reject gate for flawed items — cutting manual inspection time significantly.

Specifications & Circuit Type

Vision Hardware

Industrial GigE camera with ring-light illumination

Compute

NVIDIA Jetson Nano running an optimized YOLO-based model

Actuation

Pneumatic reject gate driven via relay from a PLC

Circuit Type

Sensor-trigger + relay-driven actuation circuit, opto-isolated from the PLC control loop

Throughput

Up to 2,400 units/hour at line speed

Reporting

Live dashboard with defect-type breakdown and shift reports

Technologies Used

Jetson Nano YOLOv8 OpenCV PLC Integration Python Industrial Camera Relay Control

Key Features

Real-time defect classification
Automated reject-gate actuation
Defect-type analytics dashboard
PLC-integrated control loop
Shift-based reporting
Configurable detection thresholds

Interested in a Similar Project?

ARC XENYX develops custom engineering, IoT, and automation solutions tailored to your business needs.

Start Your Project
Chat with us