Sensors Mag

The Power of Vision-Simplified

October 24, 2008 By: Brian Schriver


Brian Schriver

For complex inspection and part-placement applications, the vision system—and all the expenses required to support it—may be the only way to go. But for those less-demanding vision tasks, there's a solution that gets the job done without breaking the bank.

The Challenge

Machine vision suppliers have made great strides over the past 20 years, with significant improvements in the functionality and application of vision systems. These systems can be used for a tremendous variety of inspection and part-placement applications. Further advancing the technology, systems integrators have added significant value to the vision-based solutions they've created to meet industrial automation challenges. But for all the incredible increases in computing power and application capability, one challenge of vision systems remained unanswered for many years and has only recently been addressed—the overwhelming complexity of installing and maintaining these systems.

Vision systems can solve a number of applications, but they almost always require the services of a capable systems integrator. The complexity of a vision system is a result of many aspects, including the complexity of programming, multitude of camera and processor options, and variety of lens and lighting accessories. The options and flexibility give a knowledgeable vision user great power and capability, but they also put the application of vision beyond the reach of a typical end user.

The Solution

The good news is that many vision and automation suppliers have recognized this challenge and begun to introduce products that are easier to use and maintain and deliver some of the power of vision in a much more user friendly format. These products are called vision sensors. The devices are much easier to configure than traditional vision systems and typically integrate the imaging, processing, lens, and lighting in a small, rugged package, with a discrete pass/fail output. Often in sensor-like enclosures, these products also operate in a sensor-like fashion—the system provides power, tells the sensor when to take a picture, and reads a discrete output to determine if the sensor saw a good part or a bad part. The configuration software offers the user the opportunity to change a few simple parameters, teach the desired patter (a good part), and confirm that the application is running properly. Programming a vision sensor often takes less than 5 min., a far cry from the hours of scripting in the programming languages of traditional vision systems.

Vision sensors are ideal for simple inspection applications. The applications often fall into the "irritant" category—big enough to cause headaches for the manufacturing engineer but too small to justify the cost and complexity of a traditional vision system. These applications could ideally be solved by a simple photoelectric or inductive proximity sensor but prove a bit too complex for those technologies.

With a vision sensor, a manufacturing engineer can perform many inspection applications. The devices use traditional vision algorithms to identify and match patterns (or contour shapes), count pixels, or determine contrast. Pattern matching can be used to verify that small or oddly shaped components are present. It can also be used to reject damaged parts by comparing product images with known good parts. Pixel counters (sometimes call brightness detectors) can determine if a required metal component is on an otherwise plastic part; confirm that tape or plastic wrap is present on a shiny box; or check if a label has been printed on a product. Contrast is ideally suited to confirm that print is present—for example, checking to ensure a date code has been printed on a product. Vision sensors allow the user to easily identify bad products and therefore improve overall quality without dramatically increasing the complexity of the manufacturing line.

The simplicity of vision sensors is also beneficial when considering the longer-term maintenance of an application. If a user has a problem with a traditional vision system, it would typically require contacting the systems integrator who installed the system to fix it. These support calls are often expensive, even more so if the integrator has to travel. Because vision sensors are simple, a company can train its staff in the plant to setup and configure the sensors. This allows line engineers and maintenance workers to make adjustments to the application and sensor if issues come up or the inspected products change. This significantly reduces maintenance costs and downtime, compared with traditional vision systems.

Perhaps the most exciting aspect of a vision sensor to the end user is the cost. Sensing devices cost a fraction of what the hardware of a traditional vision system is priced at. That said, there is an even larger savings when you factor in the engineering costs of a traditional vision system. Vision sensors represent a substantial value when considering the bottom line.

In Closing

Vision sensors are economical alternatives to traditional visions systems for simple inspection applications. They offer simple configuration software, allowing for quick setup and easy maintenance. They also deliver lower hardware, engineering, and maintenance costs than traditional vision systems. Vision sensors use the power of vision technology to deliver application solutions in a format that can be easily applied by all users in the industrial automation sector.


Add Comment




IIoT University


Deep Learning for Vision Using CNNs and Caffe: A Hands-on Tutorial – 9/22/16 – Cambridge, Mass


IDE






Sensors 2017 Call for Speakers


Sensors Midwest


Advertise


Subscribe



Twitter Feed

Find It Fix It Forum

Sensors invites you to join the Findit-Fixit Forum, where you can get answers to your sensing questions—concerning technologies, products, methods, applications, and services--and also offer help to your fellow engineers. The Forum covers all kinds of topics, from the basics to the extraordinary.

Join the discussion!


© Copyright 2016 Questex, LLC. All Rights Reserved. Sensorsmag. Privacy Policy | Terms of Use

If you are having technical difficulties or considerations, please contact the webmaster.