Case Study

AI Vision-Based Child Part Presence Verification and PLC Interlocking System for Error-Proof Automotive Assembly

🏭 Industry

Automotive Manufacturing – Assembly Line Poka-Yoke Inspection

🧩 Problem Statement

During manual assembly operations, operators are required to install several child components onto the main assembly before the next manufacturing process begins. Missing or incorrectly assembled child parts can lead to serious quality issues, rework, warranty claims, and production losses.

Traditional inspection relied on operator confirmation, making the process vulnerable to:

Missing child parts
Wrong assembly sequence
Human inspection errors
Machine operation without complete assembly
Lack of digital traceability

The customer required an intelligent vision-based solution that verifies the presence of the child part before allowing the PLC to start the next operation.

🎯 Project Objective

Detect the presence of the required child part.
Prevent machine operation if the child part is missing.
Provide real-time visual feedback to the operator.
Display warning messages on both PLC HMI and IoT Vision Screen.
Send OK/NOK status to PLC.
Ensure 100% mistake-proof assembly.
Store inspection results for production traceability.

🛠️ Solution Overview

We implemented an AI Vision-Based Child Part Detection System integrated with the assembly station PLC and IoT monitoring software.

System Flow

Main assembly reaches the workstation.
PLC initiates the inspection cycle.
Industrial camera captures the assembly image.
AI model analyzes the predefined inspection region.
Child part presence is verified.
If the child part is detected:
- Green detection box is displayed.
- “Part OK / Part Detected” message appears.
- OK signal is sent to PLC.
- PLC starts the next assembly operation.
If the child part is absent:
- Red warning box is displayed.
- “Wrong/Absent Part” message appears.
- Alarm is shown on PLC HMI.
- Alarm is displayed on IoT screen.
- PLC blocks the next process until the correct part is assembled.
Inspection images, timestamps, operator details, and cycle status are stored in the database.

🧠 Technical Approach

AI-based Object Detection
Region of Interest (ROI) Verification
Presence/Absence Inspection Logic
Real-Time PLC Communication
Industrial Camera Image Processing
Automatic Inspection Logging
Machine Interlocking Logic
Poka-Yoke Error Prevention

Key Highlights

100% automatic child part verification
Real-time AI detection
Instant OK/NOK decision
Green/Red visual indication
PLC interlock for missing parts
Dual warning on PLC HMI and IoT screen
Automatic cycle traceability
Fast inspection suitable for production cycle time
Scalable for multiple child parts

🔌 System Architecture

Industrial Camera
LED Industrial Lighting
AI Vision Processing System (Industrial PC / NVIDIA Jetson)
PLC Interface (Ethernet/IP / Modbus TCP / Profinet / Digital I/O)
PLC HMI
IoT Monitoring Software
SQL Database
Manufacturing Execution System (Optional)

📊 Results & Benefits

✅ 100% child part presence verification

✅ Eliminated missing-part assembly defects

✅ Automatic PLC machine interlocking

✅ Dual warning on PLC HMI and IoT screen

✅ Reduced operator dependency

✅ Improved assembly quality

✅ Zero machine cycle with missing child parts

✅ Complete inspection traceability

✅ Reduced rework and customer complaints

✅ Production-ready AI Poka-Yoke solution

🏁 Conclusion

The AI Vision-Based Child Part Presence Verification System provides a robust Poka-Yoke solution for automotive assembly lines by ensuring that every mandatory child component is correctly assembled before the production cycle proceeds. Through seamless integration with industrial cameras, AI vision software, PLCs, and HMI systems, the solution prevents missing-part defects, enhances product quality, and enables complete digital traceability of every production cycle.

Download_Case_Study_pdf

AI Vision-Based 360° Differential Housing Inspection System for Chamfer, Dent, QR Code, and Cross Mark Verification

🏭 Industry

Automotive Manufacturing – Differential Housing Quality Inspection

🧩 Problem Statement

Differential housing is a critical drivetrain component that requires inspection before moving to the next manufacturing stage. Traditionally, operators visually inspect several quality parameters, including chamfer quality, dents, QR codes, and identification marks.

Manual inspection created several challenges:

Human inspection errors
Missed chamfer defects
Surface dents escaping detection
Incorrect or unreadable QR codes
Missing identification (cross) marks
Lack of inspection traceability
Increased rejection and customer complaints

The customer required a fully automated AI vision inspection system capable of inspecting every housing with high accuracy during production.

🎯 Project Objective

Detect chamfer on all required holes.
Verify chamfer quality and completeness.
Detect casting dents and surface defects.
Read and validate QR/Data Matrix codes.
Verify presence of identification (cross) mark.
Perform complete 360° inspection while the component rotates.
Send OK/NOK signal to PLC.
Store inspection images and results for traceability.

🛠️ Solution Overview

We developed an AI Vision-Based Multi-Feature Inspection System integrated with the production line and PLC for real-time quality verification.

System Flow

Differential housing reaches the inspection station.
Rotary fixture rotates the component.
Industrial camera captures images from multiple angles.
AI model simultaneously performs:
- Chamfer Inspection
- Dent Detection
- QR Code Reading
- Cross Mark Detection
Inspection results are combined.
If all parameters pass:
- PLC receives OK signal.
- Component proceeds to the next operation.
If any inspection fails:
- PLC receives NOK signal.
- Component is rejected for operator review.
Images, inspection results, and timestamps are stored in the database for complete traceability.

🧠 Technical Approach

Deep Learning–based Object Detection
AI Surface Defect Detection
QR/Data Matrix Decoding
ROI-Based Inspection
Multi-Angle Image Processing
360° Rotary Inspection
PLC Communication
Database-Based Traceability

Key Highlights

Continuous 360° inspection
Simultaneous multi-feature inspection
High-speed industrial processing
Robust under varying lighting conditions
Production-ready AI models
Automatic image logging
Easy integration with MES/ERP systems

🔌 System Architecture

Industrial Area Scan Camera
LED Industrial Lighting
Rotary Inspection Fixture
AI Vision Processing System (Industrial PC / NVIDIA Jetson)
PLC Interface (Ethernet/IP / Modbus TCP / Profinet)
SQL Database
Manufacturing Execution System (Optional)

📊 Results & Benefits

✅ 100% automated quality inspection

✅ Accurate chamfer verification

✅ Reliable dent detection

✅ QR/Data Matrix verification

✅ Cross mark detection

✅ Complete 360° component inspection

✅ Reduced manual inspection effort

✅ Improved product quality

✅ Full inspection traceability

✅ Faster production cycle

✅ Reduced customer complaints and rework

🏁 Conclusion

The AI Vision-Based Multi-Feature Inspection System automates the inspection of differential housings by combining chamfer verification, dent detection, QR code validation, and cross mark detection into a single intelligent inspection station. Through continuous 360° image acquisition and AI-powered analysis, the system delivers high-speed, reliable quality inspection while seamlessly integrating with PLCs and manufacturing systems. The solution significantly improves product quality, eliminates human inspection errors, and provides complete digital traceability for every inspected component.

Download_Case_Study_pdf

AI Vision-Based Circlip Presence Detection for Error-Proof Automotive Assembly

🏭 Industry

Automotive Manufacturing – Steering Gear / Assembly Line Poka-Yoke

🧩 Problem Statement

In steering gear assembly, a circlip is a critical retaining component that secures assembled parts in position. If the circlip is missing, improperly seated, or incorrectly assembled, the product can fail during operation, resulting in costly rework, warranty claims, and potential safety risks.

Previously, circlip verification relied on manual inspection, which was:

Operator dependent
Time-consuming
Prone to human error
Difficult to trace
Unable to prevent missing circlips before the next process

The customer required an automated vision system capable of verifying circlip presence and interlocking the assembly station with the PLC.

🎯 Project Objective

Detect the presence of the circlip before the assembly cycle is completed.
Prevent the next machine operation if the circlip is missing.
Provide instant visual feedback to the operator.
Send OK/NOK status to the PLC.
Display inspection status on the Vision Screen and PLC HMI.
Store inspection results for complete production traceability.

🛠️ Solution Overview

Wayzon Technology Services Pvt. Ltd. developed an AI Vision-Based Circlip Detection System integrated with the PLC and assembly station.

System Flow

Operator assembles the component.
PLC triggers the vision inspection cycle.
Industrial camera captures the circlip region.
AI vision algorithm inspects the predefined ROI.
The system verifies whether the circlip is correctly present.
If the circlip is detected:
- Vision screen displays CIRCLIP DETECTED.
- Green OK indication is shown.
- OK signal is transmitted to the PLC.
- PLC allows the next assembly operation.
If the circlip is missing:
- Vision system displays CIRCLIP NOT DETECTED.
- Red warning appears on the Vision Screen.
- PLC blocks the machine cycle.
- Operator must install the circlip before continuing.
Inspection status, timestamps, and production data are stored in the database for traceability.

🧠 Technical Approach

AI-Based Object Detection
ROI-Based Presence Verification
Real-Time Industrial Image Processing
PLC Communication
Automatic Machine Interlocking
Production Traceability
Industrial Lighting Compensation
High-Speed Inspection Logic

Key Highlights

100% automatic circlip verification
Real-time AI inspection
PLC interlocking for missing circlips
Green/Red visual indication
Automatic cycle validation
Industrial-grade reliability
Complete inspection traceability
Fast processing suitable for assembly line cycle time

🔌 System Architecture

Industrial Camera
LED Industrial Lighting
AI Vision Processing System (Industrial PC / NVIDIA Jetson)
PLC Interface (Ethernet/IP / Modbus TCP / Profinet / Digital I/O)
PLC HMI
Vision Monitoring Software
SQL Database
Manufacturing Execution System (Optional)

📊 Results & Benefits

✅ 100% automatic circlip presence verification

✅ Eliminated manual inspection

✅ Prevented missing circlip assembly defects

✅ Automatic PLC machine interlocking

✅ Real-time operator guidance

✅ Improved assembly quality

✅ Reduced rework and warranty claims

✅ Complete digital traceability

✅ Faster inspection cycle

✅ Industry 4.0-ready solution

Technologies Used

AI & Deep Learning Vision
Industrial Camera
PLC Integration
Industrial PC / NVIDIA Jetson
Python & OpenCV
SQL Database
Industrial IoT Dashboard
Industry 4.0 Traceability

🏁 Conclusion

The AI Vision-Based Circlip Presence Verification System developed by Wayzon Technology Services Pvt. Ltd. provides a reliable Poka-Yoke solution for automotive assembly lines. By combining AI-powered vision inspection with PLC interlocking, the system ensures that every component is assembled with the required circlip before the production cycle advances. This intelligent inspection solution minimizes human error, improves product quality, enhances traceability, and supports zero-defect manufacturing.

Download_Case_Study_pdf

IoT-Based Data Acquisition and Traceability System for Leak Test SPM

Client Overview

Sharda Motors Industries Ltd is a leading manufacturer of automotive components, specializing in exhaust systems, suspension systems, and body structures. To enhance their quality control and traceability processes, they required an advanced data acquisition and traceability solution for their Leak Test Special Purpose Machines (SPMs).

Project Summary

The IoT-based Data Acquisition and Traceability System was deployed at Sharda Motors Industries Ltd’s Pune Plant 1 & 2 and Nashik Plant 1 & 2. The system integrates with leak test machines (such as ATEQ) and marking machines (such as Marksman or Automator) to collect, store, and manage leak test data efficiently.

Challenges Faced

Lack of real-time traceability of leak test results.
Manual data logging errors leading to inconsistencies in quality reports.
Inability to link test results with unique part identification.
Repeated leak testing causing inefficiencies in production.
Delayed reporting affecting production and quality control decision-making.

Solution Implemented

The IoT-based Data Acquisition and Traceability System was implemented to overcome these challenges by enabling:

1. Automated Data Collection

Captures leak test data in LPM (Liters Per Minute) directly from leak test machines like ATEQ.
Links test data with unique part identification by integrating with Marksman and Automator marking machines.
Provides unique QR codes with details including leak value, date, time, shift, company logo, ensuring easy traceability.
Barcode printers are used to print QR code labels, which are then pasted onto the respective parts.
Scanners are provided to ensure Poka-Yoke validation, confirming that the QR code is affixed to the correct part before moving forward.

2. Cloud-Based Data Traceability

All leak test data and generated QR codes are stored securely on the cloud.
Data can be accessed anytime, from anywhere, via a dedicated URL.
Enables real-time tracking of production, quality, and maintenance data.

3. Automated Reports & Notifications

Daily production analysis reports are automatically emailed to registered recipients.
Quality, production, and maintenance reports are accessible online and shared via email or mobile notifications.
Maintenance alarms generated by the PLC are sent in real-time to the concerned department, reducing downtime and improving efficiency.

Key Business Impact

1. Increased Production Efficiency

At Lumax, the implementation of the IoT system resulted in a 25% increase in production.

2. Reduction in Repeated Leak Testing

At Sharda Motors, repeated leak testing was reduced by 30%, improving overall efficiency and reducing waste. Production is increased by 20% as real time data is shown on cloud system for monitoring purpose. Machine ideal state is reduced from 2.3 hrs to 15 minutes .

3. Enhanced Traceability & Compliance

Each part undergoes complete traceability, ensuring adherence to quality standards and industry regulations.

4. Real-Time Data Access & Decision Making

With cloud-based storage, management teams can access reports anytime and make data-driven decisions.
Mobile and email notifications ensure instant alerts for any anomalies in production or maintenance requirements.

Conclusion

The IoT-based Data Acquisition and Traceability System has significantly improved operational efficiency, traceability, and production quality at Sharda Motors Industries Ltd. By leveraging real-time data collection, cloud storage, and automated reporting, the company has achieved greater accuracy in leak testing while reducing downtime and optimizing production workflows.

This successful implementation highlights the power of IoT in industrial automation and serves as a model for other automotive manufacturing companies looking to improve quality control and production efficiency through smart data acquisition and traceability solutions.

Download_Leak_Test_SPMCase_Study_pdf

Vision-Based Quality Inspection and Traceability System

Client Overview

Sharda Motors Industries Ltd is a leading automotive components manufacturer, known for its high standards in quality control and production efficiency. To enhance process automation and traceability, the company required an advanced vision-based system for ensuring circlip presence verification during assembly.

Project Summary

The Vision-Based Quality Inspection and Traceability System was implemented at Sharda Motors Industries Ltd to ensure accurate circlip pressing during production. This system integrates with machine vision technology and IoT-based data acquisition to enhance precision, reduce errors, and improve traceability.

Challenges Faced

Manual inspection inefficiencies leading to incorrect circlip placement.
Production delays due to idle machine time caused by rework.
Lack of traceability of individual parts in the assembly line.
Quality inconsistencies affecting compliance with industry standards.

Solution Implemented

The Vision-Based Quality Inspection and Traceability System was deployed to address these challenges by enabling:

1. Automated Circlip Presence Detection

A vision system detects whether the circlip is properly placed on the part.
If the circlip is detected, the system sends a forward command to the cylinder, allowing the circlip to be clipped onto the part.
If the circlip is missing, the system prevents further machine operation, ensuring quality compliance.

2. QR Code Generation and Scanning for Traceability

Once the circlip is successfully clipped, a unique QR code is generated containing:
- Unique ID
- Date & Time
- Part Details
- Vendor Details
- Shift Information
- Production Count
The QR code is printed and pasted on the part to ensure traceability.
Before the part is unclamped, the QR code is scanned to confirm all details are correctly registered in the system.
Once validated, the machine releases the part and is ready for the next cycle.

3. Data Acquisition and Cloud-Based Traceability

All inspection data, QR codes, and production logs are stored securely on the cloud.
Data is accessible in real time via a dedicated URL for production monitoring.
Daily production analysis reports are emailed to registered recipients.
Quality reports, production reports, and maintenance alerts are shared via email and mobile notifications.

Key Business Impact

1. Increased Production Efficiency

Machine idle time was reduced by 1 hour and 50 minutes, leading to a significant boost in production output.

2. Reduction in Rework & Quality Issues

The vision-based system eliminated manual inspection errors, ensuring 100% accurate circlip placement.

3. Enhanced Traceability & Compliance

Each part is now uniquely identified with a QR code, ensuring full traceability throughout the supply chain.

4. Real-Time Data Access & Decision Making

Management teams can access reports anytime from any location, enabling data-driven decisions.
Automated alerts improve response time to production or maintenance issues.

Conclusion

The Vision-Based Quality Inspection and Traceability System at Sharda Motors Industries Ltd has significantly improved production efficiency, traceability, and quality assurance. By leveraging machine vision technology, IoT, and cloud-based data management, Sharda Motors has successfully optimized machine uptime, reduced errors, and enhanced production workflows.

This successful implementation highlights how vision-based automation can revolutionize quality control in automotive manufacturing by eliminating manual errors and increasing process efficiency.

Download_vision-based_Case_Study_pdf

Vision-Based Arrow Direction Detection on Muffler for Automated Welding

🏭 Industry

Automotive Manufacturing – Exhaust / Muffler Welding Line

🧩 Problem Statement

In the muffler welding process, correct orientation of the muffler is critical.
Each muffler has a directional arrow marking, and welding must start only if the arrow direction is correct.

Earlier, this verification was:

Manual
Error-prone
Causing incorrect welds and rework

The client required an automated, reliable, and PLC-integrated vision solution to ensure zero welding on wrongly oriented parts.

🎯 Project Objective

Detect arrow marking on the muffler using a camera
Identify arrow direction (Left / Right / Correct orientation)
Send OK / NOT OK signal to PLC
Allow welding operation only when orientation is correct

🛠️ Solution Overview

We implemented a Vision-Based Arrow Detection System with real-time PLC communication.

System Flow:

Camera captures muffler image before welding
Vision algorithm detects arrow marking
Arrow direction is calculated
If direction matches reference:
- PLC receives OK signal
- Welding operation starts
If direction is wrong:
- PLC blocks welding
- Operator intervention required

🧠 Technical Approach

Edge-based arrow detection
Direction vector analysis
ROI-based processing for high speed
Noise-resistant logic suitable for welding environment

Key Highlights:

Works under industrial lighting
Robust against dust and surface reflections
Fast decision-making suitable for cycle-time constraints

🔌 System Architecture

Industrial Camera mounted near welding station
Vision Processing System (Industrial PC / Jetson)
PLC Interface via Digital I/O / Ethernet
Welding Machine Interlock

📊 Results & Benefits

✅ 100% prevention of wrong-orientation welding
✅ Zero manual inspection required
✅ Improved weld quality and consistency
✅ Reduced rework and scrap
✅ Seamless PLC integration
✅ Industry-ready and scalable solution

🏁 Conclusion

This project demonstrates how Vision + PLC integration can eliminate human error and enable fully automated, decision-driven welding operations.
The solution is production-proven, scalable, and adaptable to other orientation or marking-based inspection tasks.

Download PDF Version