Relimetrics provides shop floor ready AI-based quality automation solutions that enable companies to digitize their quality assurance (QA) inspections and process control in manufacturing and assembly. Relimetrics solutions are designed for high production variability manufacturing environments with real-time inspection needs , can be quickly installed on the shop floor and provide highly accurate anomaly detection. Relimetrics is an industrial-grade, flexible and scalable image inspection solution, which can be managed directly by our customers. Interface modules enable connection to any manufacturing control and resource planning system, enabling a closed loop manufacturing process to be delivered to any production facility. Relimetrics software can run both in the cloud and at the edge, have an open API enabling easy addition of new functionality, and consists of 3 modules: Training Module [aka PIXEMETRICS] enables customers to develop in-house deep learning models and extend the detection capabilities of these models to new parts and configurations. Quality Audit Module runs the deep learning models trained by PIXEMETRICS and digitizes QA inspections in-line, creating full traceability of quality across products inspected down to individual product components. Process Control Module provides real time process control feedback to improve any manufacturing process by correlating machine and process data with digitized quality data.
Offerings
Offering
Relimetrics and Intel® are working together to bring you the Relimetrics AI-Based QA automation system for electronics assembly. Powered by Intel® architecture, the solution analyzes an assembled product as it comes through a conveyor belt for quality inspection. High-definition cameras perform visual inspection of the assembled product. The high-resolution video stream is then transferred to an embedded or attached IT system for analysis by Relimetrics video analytics application running at the edge. The application uses Intel® CPUs and VPUs to perform advanced computer vision and machine learning at the edge, fully digitizing the QA inspection of the product. The results are then against the bill of materials the manufacturing execution system defined during the build process. If an anomaly is detected, the system will send an alert detailing the identified issue. Staff can pull the defective product and fix the issue before shipping it to the customer. Further data analytics and model re-training can then be performed at the edge or in the cloud based on customer needs. By implementing this solution, at IT electronics manufacturer producing 30,000 servers per month can reduce the cost of rework by half a million Euros per manufacturing line per year. What's more, the solution slashes the time to pinpoint the root case of quality issues from days to minutes.