As the name suggests, “machine vision” is a term referring to a computer’s ability to “see” – at least in a technical sense.
Devices equipped with machine vision capabilities deploy one or more video cameras and the recorded data goes directly to either a computer or a robotic controller. Based on that information, the computer can then perform a wide range of different tasks that would typically require human intervention.
A closely related concept called computer vision is used in a large number of ways, including for things like signature identification, object recognition, currency inspection and more. But what really makes machine vision important has to do with the implications it has already brought with it to the world of industrial robotics.
Machine Vision and Industrial Robotics: The Perfect Pairing
More often than not, machine vision is used for the purposes of robot guidance – that is, it empowers an industrial robot’s ability to understand its own location within its frame of reference, which is information it can then use to chart a path towards some goal location.
For the sake of example, let’s say an industrial robot is on one side of a room and an important tool is on a table about 15 feet away. The “end goal” is that this robot will cross the room and use a mechanical arm to “grab” the tool.
Using machine vision, that industrial robot will first establish its position within the room using the associated video cameras and other hardware. It will then locate the tool, or at the very least the handle that it can grab onto. Then, it can use spatial information to better understand the position and orientation of the tool, at which point this data will be transferred over to the industrial robot’s coordinate system.
Armed with that information, the industrial robot can guide itself to the ideal position, adjust its own orientation to make it possible to pick up the tool, and do so. But the most impressive feat of all is that this entire chain of events can be executed in a matter of seconds.
The Broader Implications of Machine Vision
This may be an admittedly simplified example, but the larger implications of machine vision in the world of industrial robotics is enormous. First and foremost, advancements in machine vision have dramatically increased the demand for industrial automation and as these systems get better and more accurate, it allows industrial robots to do a larger share of the work in places like factories.
But the real benefit to businesses is that machine vision in this context allows for fast, high quality production control. Industrial robots with machine vision can quickly spot defective parts far faster than a human could, thus allowing organizational leaders to take the appropriate action. The type of visual information obtained through machine vision can even be used to prompt predictive maintenance plans, allowing people to stop small problems with equipment today before they turn into much more costly periods of unexpected downtime tomorrow.
Of course, getting to this point isn’t without its fair share of challenges. Experts at FANUC agree that proper image calibration is of paramount importance, as if mistakes are made during this process the industrial robot may not “understand” where it is relative to a particular object, thus creating a potentially problematic situation. Likewise, elements of an environment like lighting are crucial for helping industrial robots “see” accurately.
But still, the increased accuracy of machine vision guidance systems has created increased robotic accuracy over the last five years alone. What the public once perceived as something too complicated for its own good has now turned into a race for manufacturers and industrial professionals to capitalize on new opportunities to use it, which is a benefit for businesses and their consumers alike.
Contact us today!
Read more about robotics and industry trends from our weekly blog posts
Find out what to expect when working with Pioneer
Follow our LinkedIn page for company updates and exciting news
