Automated Inline Inspection and Sorting of Labeled Products Using Machine Vision and PLC Control

Document Type

Conference Proceeding

Publication Date

6-24-2026

Department

Department of Mechanical and Aerospace Engineering; Department of Applied Computing

Abstract

We present a compact mechatronic cell for inline inspection and sorting that couples RGB/depth vision on a Raspberry Pi with deterministic actuation on a CompactLogix PLC. The Pi decodes barcodes, validates semantic fields (SKU, batch, expiry), and transmits a binary decision to the PLC, which drives pneumatic diverters. On 50 bottle passes, the prototype achieved 97% correct detection for valid products and 100% correct rejection of unlabeled items. Novelty lies in the clean SBC-PLC split that keeps perception flexible while preserving industrial timing and safety, enabling low-cost retrofits. Limitations: accuracy is sensitive to label glare and occlusion; actuation timing is currently time-based rather than sensor-triggered. Scalability: the architecture extends to multi-lane lines and SKU changes via software-only updates, with straightforward hooks for SCADA/MES logging.

Publication Title

2025 International Conference on Engineering and Computing Technologies (EngiTek)

ISBN

979-8-3315-6047-8

Share

COinS