Bottle Exterior Defect Detection

csdetb-icono

Customer

A major Japanese multinational cosmetics and personal care company.

Project Details

  • Duration: 4 Months

Technologies:

  • AI Vision
  • OpenCV
  • MongoDB
  • RabbitMQ
  • Python
problem-img

Problem

  • Manual inspection struggles to detect fine exterior defects such as scratches, cracks, dents, and dirt.
  • Lack of consistent and data-driven quality monitoring across multiple factories.
  • High risk of defective bottles reaching customers, causing brand damage and increased product recalls.
solution-img

Solution

  • We developed an AI-powered vision inspection model to detect bottle exterior defects in real time.
  • High-resolution cameras integrated with edge AI to spot cracks, dents, scratches, and foreign particles.
  • The model ensures manual adjustment of thresholds for different defect types (e.g., scratches, Dents, Foreign Particles, etc.).
  • With the deployment of this solution, 10,000+ bottles per station are inspected on a daily basis.

97+%

Accuracy

150%

Increased efficiency in Inspection

Key Features

cskeyfb-img1

Real-Time Monitoring

cskeyfb-img2

Auto-Training Pipeline of New Productions

cskeyfb-img3

Live Comparision: Actual Production vs Planned Production

Insights on innovation

Stay updated with the trending and most impactful tech insights. Check out the expert analyses, real-world applications, and forward-thinking ideas that shape the future of AI Computer Vision and innovation.

June 26, 2026 - 4 minutes to read

Do It Safe, Then Prove It: Insurance-Grade Footage for Yard Safety Compliance

Three months after the incident, the assessor sits across from Meera, the operations manager, and slides a form over the desk. She is ready for questions about helmets, about whether the supervisor was present, about protocol. He asks none of them. He asks one thing instead: show me the footage. Meera’s yard runs safety well. […]

Read More

Ruchir Kakkad

CEO & Co-founder

June 24, 2026 - 4 minutes to read

22 vs 30 Moves an Hour: Finding the Hidden 1% Gains in Terminal Productivity

Deepak runs operations at a mid-size terminal, and one number has bothered him for a month. Two ship-to-shore cranes, side by side. Same model, same year, same maintenance schedule. Crews rotate between them every shift. One averages 22 moves an hour. The other averages 30. He sends a maintenance team over the slow one. They […]

Read More

Ruchir Kakkad

CEO & Co-founder

June 23, 2026 - 4 minutes to read

Empties Are Inventory, Not Furniture: Smart Empty Container Yard and CFS Management

Pravin has run the empty depot for nine years, and he will tell you he knows every box in it. Then he walks you to the far corner, points at a container furred with dust, a small spider web in the corner casting, and goes quiet. It came in months ago. Nobody remembers the booking. […]

Read More

Ruchir Kakkad

CEO & Co-founder

Whatsapp Img