Full autonomy is the ultimate goal of industrial automation. Vision technology is inarguably a crucial ingredient in making this goal a reality. The evaluation of visual cues will enable a robot to attain a sufficient understanding of its environment and to execute innumerable tasks without intervention from a human operator.

So why do we see so little proliferation of machine vision in contemporary automation systems? The simple answer is that machine vision is complex and comes with significant risks for system integrators, unless they employ a staff of dedicated engineers, providing services such as connecting hardware, labeling training data, deploying AI-models, tuning parameters, etc. 

We believe that most of these services that drive the cost of machine vision in industrial automation can be made superfluous or also be automated.