Md Rahinur Rahman
Lead AI Systems Engineer
Leads the design and development of industrial automation for Saturn R&D platforms.
Future of Fabric Inspection
FABINSFabric Inspection Automation
Detects every defectMeasures roll dimensionsCertifies fabric quality
Automate Fabric Inspection with FABINS.
AI defect detectionMillimetre measurementFour-Point scoringAutomatic reportsEncoder-locked captureRetrofit installation
The problemTHE BOTTLENECK IN
THE BOTTLENECK IN
EVERY FABRIC ROLL
Bangladesh RMG is a national export pillar, yet quality control at the frame is still one inspector, one pair of eyes, one tally sheet — while competing manufacturing hubs have already automated it.
Manual, Person-Dependent
Grading still relies on an inspector's eye at the frame — not on installed machine capacity.
Fatigue & Monotony
Long shifts watching moving cloth degrade attention — a physiological limit, not a discipline issue.
Inconsistent Quality
Two inspectors grade the same roll differently — human error is unavoidable, not occasional.
Almost No Usable Data
A tally sheet, not a defect map — nothing to trace faults back to a machine or batch.
The StrategyRETROFIT, DON'T REPLACE:
RETROFIT, DON'T REPLACE:
UPGRADE YOUR EXISTING MACHINES
Importing complete automated inspection machines from abroad requires millions in capital expenditure and disruptive factory floor redesign. FABINS upgrades the inspection machines already sitting in Bangladeshi mills.
Dimension
Conventional Machine
FABINS Retrofit
Approach
Replace
Replace the complete inspection machine outright
Upgrade
Upgrade the existing inspection machine as a retrofit
Investment
High Capital
New machine purchased outright at high foreign cost
Low-Cost Retrofit
Low-cost retrofit integration into existing mill hardware
Installation
Long Downtime
Long changeover and extended factory commissioning
Fast Deployment
Fast deployment with minimal production disruption
Support & Fit
Vendor-Dependent
Overseas service with high barrier for mid-sized mills
Locally Engineered
Locally engineered & supported for local fabrics and specs
The systemFROM ROLLING FABRIC
FROM ROLLING FABRIC
TO INSTANT REPORT.
Synchronized with real fabric motion to deliver millimeter-accurate defect detection at full production speed.
Step 01
Phase 1 of 6Camera Captures
High-speed line-scan camera captures the full fabric width as it moves.
FABINS Inspection Pipeline01 / 06
Step 02
Phase 2 of 6Image Acquisition
Rotary encoder triggers every image line from actual fabric travel for precise sync.
FABINS Inspection Pipeline02 / 06
Step 03
Phase 3 of 6AI Defect Detection
Deep learning vision model resolves sub-millimeter defects on moving fabric in real time.
FABINS Inspection Pipeline03 / 06
Step 04
Phase 4 of 6Classify & Measure (mm)
Distinguishes holes, joints, oil spots, and measures exact physical size in millimeters.
FABINS Inspection Pipeline04 / 06
Step 05
Phase 5 of 6Four-Point Grading
Applies Four-Point penalty rules automatically, including fixed hole penalty overrides.
FABINS Inspection Pipeline05 / 06
Step 06
Phase 6 of 6Report & Dashboard
Issues mill's own official inspection report format, signed and ready for sign-off.
FABINS Inspection Pipeline06 / 06
Built on catalogue industrial hardware
8192px Line Rate
Line-Scan Camera
8192px Industrial Color Sensor
Industrial high-resolution line-scan camera capturing the full fabric width with sub-millimeter optical clarity per line.
Rotary Hardware Pulse
Lens & Encoder
Encoder-Triggered Sync
Calibrated optics paired with a rotary hardware encoder, locking capture geometry to actual fabric speed regardless of conveyor variance.
Uniform High-CRI LED
Lighting & Conveyor
Controlled Illumination Rig
Custom-designed uniform lighting assembly over a continuous-motion fabric transport rig, eliminating shadow artifacts.
Real-Time Edge Pipeline
AI Software & Dashboard
Browser Operator Interface
Detection, grading, and reporting pipeline with an intuitive browser-based operator dashboard for live feeds and report downloads.
Defects & Four-Point GradingWHAT IS A FABRIC DEFECT &
WHAT IS A FABRIC DEFECT &
HOW IS IT SCORED?
Fabric faults are detected in real-time and automatically scored using the industry-standard Four-Point System.
1. The ASTM D5430 Four-Point Penalty Rules
1 PointRule 1
Up to 3 inches (75 mm)
Minor defect or short localized yarn imperfection.
2 PointsRule 2
3 to 6 inches (75 – 150 mm)
Medium imperfection extending across several warp/weft yarns.
3 PointsRule 3
6 to 9 inches (150 – 230 mm)
Major imperfection affecting fabric appearance or strength.
4 PointsRule 4
Over 9 inches (230 mm+) or Holes
Critical defect or any physical hole/tear regardless of size.
2. Common Fabric Defects
ASTM D5430 Defect Taxonomy
3. How Total Roll Grade is Calculated
ASTM D5430 Standard Formula
After inspecting a fabric roll, total penalty points are normalized per 100 square yards to determine if the roll passes export quality standards.
Official Four-Point Formula
Points / 100 yd² = (Total Points × 3,600) ÷ (Yards × Width Ins)
Example Roll Calculation:
Fabric Roll Spec:100 Yards Length × 60 Inches Width
Defects Found:3 Slubs (3 pts) + 1 Hole (4 pts) = 7 Pts
Formula Output:(7 × 3,600) ÷ (100 × 60) = 4.2 Pts
Acceptance Result:PASS (Grade A · Limit: 40 pts)
Standard Buyer Acceptance Thresholds
0 – 20 PtsGrade A (Pass)
21 – 40 PtsGrade B (Pass)
> 40 PtsREJECT / Cut
FABINS automates this entire calculation process instantly as the fabric unrolls.
The Innovators
INNOVATORS BEHINDFABINS
System architecture, hardware integration, computer vision and production software — built at the Saturn Textiles Ltd R&D department.
Mohammad Ninad Mahmud Nobo
Lead AI Software Engineer
Leads full-stack web development and machine learning model integration for Saturn R&D platforms.
LET'S CONNECT
Have a question, technical inquiry, or partnership request? Send a message to our R&D team below and we will respond within 1–2 business days.