Vision 360

Vision 360 is a solution that improves manufacturing inspection accuracy with image acquisition and deep learning analytics.  It takes full advantage of the advances in AI and parallel computing. This groundbreaking technology goes beyond detection of visual information by providing actional insights to improve quality, yield and cycle time.

Heretofore, machine vision systems are passive in the production process and prone to provide false or missed alerts.

Vision 360 performs the following:

• Captures rich information by the vision system—augmented with AI—becomes a real-time proactive part of the production line process
• Classifies and forecasts key KPIs—with real time clustering—provides early identification of potential problem areas and continuously steers the production to operate at its highest point of efficiency
• Integrates production lines that include process machines along with optical inspections to drive faster and more accurate decisions during the manufacturing process
• Leverages the latest advantages in GPU computing, transforms the visual inputs sub-second to provide inferences in real time, it is offered on cloud, private cloud or inhouse infrastructure

Features

• Image Acquisition – Connectors to a wide-range of optical inspection systems and camera devices
• Pre-built Vision Inspection Templates – Templatized models for rapid set-up, training and deployment into production
• Anomaly detection and Prediction – Anomaly detection and prediction using deep AI models
• Real-Time Quality Assessment – Monitor in real-time production quality with automated defect detection, production unit health scores, alerts and notifications
• Automated Defect Classification – AI models that combines live Image data with operating data for accurate defect classification and sorting
• Deep Forecasting – Deep learning AI models layered with Image AI for quality forecasting and pre-cursor detection
• Soft-sensing – Soft-sensing of critical process parameters in real-time
• Optimization – Real-time Process Optimization Advisories to run production close to golden batch conditions
• Model Management – Automated model re-training for anomaly detection, forecasts, soft-sensing, and optimisation
• MLOps, Continuous integration, continuous deployment – No resource drain for end-users or IT staffs to keep up with new releases and feature enhancements