Skip to main content

Industry

Alcoholic Beverages (Alco-Bev)

Result 1

99%+ accuracy in detecting label defects and text errors

Deployments

Kompass AI and Tigris (Modular Vision Hardware)

Result 2

Faster, more reliable in-process inspection with automated ejection and alerts


By deploying Jidoka’s AI-driven inspection systems, Diageo streamlined its label verification(LBI/Fill level) process at 300 bottles per minute, ensuring accurate text, proper orientation, and accurate label – while minimizing manual oversight.

Overview

‍Diageo is a global leader in premium spirits, producing iconic brands across more than 180 countries. Known for high-volume bottling lines, Diageo needed a robust, automated solution to ensure precise labeling – both for compliance and brand consistency – without slowing production.

‍The Opportunity

‍At Diageo, manual label inspection was error-prone and unsustainable at high speeds:

  • 300 bottles/minute demanded real-time verification of text, logos, barcodes, and label orientation.
  • Bottles with torn, scuffed, or bubbled labels could slip through, risking brand image and regulatory non-compliance.
  • Pre-printed and variable text needed accurate reading and matching against templates—human oversight alone wasn’t enough.

‍Jidoka installed Kompass and Tigris to tackle Diageo’s labeling challenges:‍

1. Real-Time OCR & Object Detection
  • High-resolution cameras captured label regions, reading critical text (brand names, disclaimers, barcodes) in milliseconds.
  • Object detection verified logo presence, correct orientation, and consistent label placement.
2. Defect Identification & Classification
  • Automated detection of label tears, trapped bubbles, and scuff marks, triggering immediate alerts.
  • Any misprint or mismatch from the approved template was flagged for rejection.
3. In-Process Inspection & Ejection
  • Jidoka’s hardware module, Tigris, seamlessly integrated with Diageo’s bottling lines, ejecting faulty bottles on the fly.
  • An andon-like alert system informed operators of recurring issues, speeding up root-cause analysis.
4. Adaptable for Multiple SKUs
  • The AI-driven approach allowed quick configuration changes for new label designs, compliance requirements, or product variations.