Limitations of traditional machine vision require factories to set up secondary manual inspections that create bottlenecks in production.

Traditional machine vision systems generally have limited functionality, require controlled environments, and produce a high rate of false calls. These limitations stem from the “golden image” approach to model training where any variance from a particular control set is deemed defective. But many manufacturing requirements allow certain variances too complex for machine vision systems to understand, leading to bottlenecks in production.

AI Platforms lead with promises of efficiency but result in long training cycles and pilot purgatory due to lack of available datasets.

AI Platforms lead with promises of efficiency but result in long training cycles and pilot purgatory due to lack of available datasets.

LIVIS is a true deep learning system built for manufacturing. The system includes proprietary AI models trained on industry- and use case- specific datasets, enabling manufacturers to deploy significantly faster than with other AI platforms. Unbound by rule-based structures, LIVIS gets smarter over time by learning which variances are acceptable and which are not. As a result, LIVIS consistently outperforms traditional machine vision technology and other AI platforms in accuracy, latency, and speed.