Stationery / Pencil Manufacturing
99% defect detection accuracy
Kompass (AI) Vision System with Multi-Camera + Laser Measurement Integration
125 parts per minute inspection speed
Overview
Hindustan Pencils Pvt. Ltd., India’s largest pencil manufacturer and a cornerstone of the Indian stationery industry, produces millions of pencils that demand consistent quality and flawless finish.
A critical stage in pencil manufacturing is the inspection of wooden slats—the foundation of every pencil. Surface defects such as cracks, knots, pin holes, bends, and dimensional inconsistencies directly impact downstream processes and can result in 2–5% defective pencils during later production stages.
Opportunity
Ensuring consistent slat quality at high production speeds presented several challenges:
- Manual inspection variability between operators affected detection reliability
- Defect escapes led to rework and downstream production inefficiencies
- Required inspection of multiple surfaces (top, bottom, sides) within a single station
- Manual dimensional verification could not reliably keep pace with high production speeds.
- Required real-time decision-making integrated with the existing PLC system
The goal was clear - eliminate manual dependency and create a high-speed, zero-down time inspection system.
Jidoka's Approach
Vision & Laser-Based Inline Inspection
A top-mounted camera captures the required regions of interest on the wooden slats for visual inspection, while a 3D laser profiler performs precise dimensional measurements. The laser system identifies critical dimensional deviations including bend, width and length variations and thickness variation, ensuring accurate verification of the slat geometry.
AI-Powered Visual Defect Detection
All captured images are processed through Jidoka’s deep-learning KOMPASS engine, enabling reliable detection of visual defects such as knots, cracks, hard wood spots and pinholes. The AI-based inspection ensures consistent identification of surface anomalies that may affect product quality.
Intelligent Defect Classification
All detected defects are automatically categorized into critical, major or minor classes based on predefined thresholds, enabling standardized and objective decision-making without manual intervention.
High-Speed Inline Integration & Automated Rejection
The inspection system is seamlessly integrated with HPPL’s existing conveyor line, capable of supporting high-speed operations of up to 450 parts per minute. Based on the inspection outcome, the system triggers eight pneumatic ejection outputs to automatically segregate defective parts without interrupting production flow.
Comprehensive Coverage & Process Reliability
With a field of view covering the entire inspection area of the wooden slat, the system ensures complete inspection of critical surfaces, enabling consistent detection accuracy and eliminating reliance on manual inspection.
PLC Integration & Automated Rejection
Image capture is triggered by Jidoka’s PLC in synchronization with the inspection cycle. The inspection results (OK/NOK) are communicated to the machine in real time, enabling immediate action. The system incorporates eight reject gates to ensure accurate segregation of defective wooden slats. Electrical interlocks enable seamless integration with the existing conveyor and station controls, while OK parts continue downstream without interruption.
KnotCrackHard woodChip