A leading tyre manufacturer aimed to streamline its inspection process for Tyre Identification Numbers (TIN). Manual methods were slow, prone to errors, and unable to scale for a daily throughput of 10,000+ tyres. By implementing an automated visual inspection system, the client sought to reduce operational costs and enhance product quality.
Operating in a highly competitive tyre manufacturing sector, the client produces thousands of tyres daily. Each tyre must have a clearly identifiable TIN for quality control, regulatory compliance, and traceability.
With an existing production management tool already in place, the client needed a robust, AI-driven inspection solution that could integrate seamlessly, reduce manual overhead, and deliver near-instant TIN recognition.
Manual Bottlenecks: Human inspectors had difficulty maintaining speed and accuracy, especially given the daily volume of 10,000 tyres. Error-Prone Process: Misread or overlooked TINs led to production slowdowns, rework, and potential compliance issues. Integration Hurdles: Any new system had to fit seamlessly into the client’s existing workflow and “Harwa” production management tool.A camera-based automated inspection system was deployed, featuring AI models capable of detecting and reading TINs in under two seconds per tyre. Once identified, results are instantly updated in the existing production management software.
Requirement Analysis: Detailed study of the tyre production line to identify optimal camera placement and system requirements. AI Model Development: Use of deep learning algorithms to accurately detect TINs under varying lighting conditions and tyre orientations. Integration: Real-time data exchange with the Harwa production management tool to ensure immediate feedback on defective or unreadable TINs.Assessment & Planning:
The team conducted on-site evaluations to determine camera types and installation points. A rollout plan was drafted to ensure minimal disruption to daily operations.Deployment & Integration:
Cameras were installed on the inspection line to capture tyre TINs. AI models, developed using PyTorch and OpenCV, were integrated into a Python/Django backend. This backend communicated with both the camera systems and the Harwa tool.Training & Optimization:
Real-world data from various tyre batches was used to refine the AI models, improving accuracy and reducing false positives. Production staff received hands-on training to manage alerts and address any flagged tyres.Continuous Monitoring & Support:
Regular performance checks ensured the system maintained sub-two-second identification speeds. Ongoing updates further optimized detection accuracy and minimized downtime. Camera Systems:
High-resolution industrial cameras calibrated for TIN visibility. AI Software:
Deep learning frameworks (PyTorch) for model training; OpenCV for image processing. Backend & Integration:
Python and Django for server-side logic; seamless data exchange with the Harwa. Frontend & Visualization:
React-based dashboard providing real-time TIN detection results and alerts. Specialized Camera Installation:
Mounted high-speed cameras at optimal angles along the production line to capture TINs consistently. Iterative AI Model Deployment:
Incrementally rolled out and tested AI models, refining them with real-time feedback from live tyre inspections. System Integration & Dashboard Development:
Deployed a React-based interface for operators, displaying tyre inspection outcomes and enabling quick interventions for unreadable TINs. Comprehensive Testing & Validation:
Conducted multiple pilot runs under different lighting and tyre orientation scenarios to confirm consistent sub-two-second detection speeds and high accuracy.Rapid TIN Identification: Achieved identification speeds under two seconds per tyre, enabling efficient processing of 10,000+ tyres daily.Reduced Operational Costs: By minimizing manual inspections, labor requirements and error-related rework were significantly lowered. Enhanced Quality Assurance: Faulty or unreadable TINs are flagged instantly, preventing defective tyres from advancing in the production line. Streamlined Workflow: Integration with the Harwa system allows immediate updates, ensuring any issues are addressed in real time and operators remain well-informed. 24/7 Automated Process: Continuous inspection eliminates the need for shift-based manual oversight, maintaining consistent productivity and accuracy.
Book an Appointment
