Manufacturing AUTOMATION

Features Factory
How vision-based optical sensors marry two technologies for improved error-proofing


March 3, 2008
By Tom Draper and Mark Sippel

Topics

When it comes to the level of quality that is acceptable for today’s manufacturers, one thing is clear. Manufacturers can only meet their customers’ requirements by making parts that are 100 per cent perfect all the time – while still reducing costs and improving productivity.

A few years ago, it was the PLC’s job to ensure quality parts. Today, the task of ensuring quality rests on sensors and vision systems to make sure products are in spec. In fact, the merger between sensors and vision systems is the next step forward in sensing-driven product quality improvement.

Sensor-driven error-proofing
Over the past few years, manufacturers have used terms such as total quality management, lean manufacturing, error-proofing, mistake proofing and Poka Yoke to describe various programs designed to drive out mistakes in the manufacturing process. Today, the most effective error-proofing programs use sensors to eliminate errors as an integral part of the manufacturing process itself. Sensors error-proof each manufacturing step to prevent out-of-spec parts from being produced in the first place by checking each step and either eliminating a part that was produced out of tolerance or stopping the process altogether.

Most discrete sensor-driven error-proofing works best when the inspection system can automatically position each manufactured part exactly the same way so it can aim the sensors to verify a specific aspect of the part, serving as an indicator that the manufacturing step was done correctly. This calls for sensors stationed all along the way to make their individual mini-inspections.

Advertisment

Vision-based inspection typically calls for expensive equipment to mimic old-fashioned in-line or final inspection method in an automated way. On the other hand, vision methodology has the advantages of being able to inspect parts in various attitudes relative to the camera and inspect more than one attribute simultaneously, including appearance, presence/absence, dimensional attitudes and positioning.

These technologies tend to be polar opposites in both complexity and capability. In the past, sensor-driven error-proofing has been limited to discrete functionality based on specific technology. This includes photoelectric sensing, proximity sensing or laser based sensing to error-proof a production or manufacturing process. This process works well in the discrete manufacturing arena, providing relatively inexpensive solutions based on application expertise. On the opposite side of the spectrum, vision typically provides more complex multi-task sensing involving many simultaneous sensing methods or algorithms that can perform sensing operations for error-proofing as discrete sensors with the addition of complex sensing requiring interconnection between sensing methods.

In terms of the technology, there are more sophisticated sensors and application techniques advancing up the curve towards vision solutions. At the same time, there is vision equipment moving down the curve towards the discrete sensor world. But instead of a crash of technologies, there is evolving a new layer of technology that combines the best from each. With the combination of both technologies and the simplified “sensor-like” approach to configuration and usage, the user can apply higher level sensing at a lower cost point, allowing these new optical sensors to be applied more readily in a true error-proofing scheme.

Simplified vision technology
These new vision-based optical sensor products bridge the gap between the two technologies by providing a simple, practical and cost-effective way to error-proof production by simultaneously checking several aspects of the product with a single device. These new devices use a simple configuration interface that in-house staff can quickly learn and use. New optical sensors with simplified configuration interfaces and multiple inspection/measurement algorithms driving multiple sensing options provide more information than a single-function smart camera or a standard discrete sensor. At the same time, they avoid the traps of complex vision systems in cost, complexity and needed expertise for achieving reliable error-proofing.

When and how should manufacturers use these new optical sensors? This new type of vision-based sensor is more like a smart sensor than a vision system. Just like a sensor, it is configured to look for certain attributes of a part or product to make sure specific aspects of the product are present, the part is configured correctly and positioning is verified. But unlike a discrete sensor, the optical sensor does not need the part to be presented exactly the same way for each inspection, thus reducing fixturing costs. And unlike a discrete sensor, vision sensors can check for multiple characteristics at the same time, justifying their cost with a higher ROI sooner by taking the place of several sensors, each of which can only check on one thing at a time. As opposed to using a more traditional sensing array, these optical sensors can also significantly reduce the complexity and cost of error-proofing while improving overall reliability. This opens up a whole new world of error-proofing that was not available before for reducing planned downtime, easier line changeovers and better accommodating flexible manufacturing.

The optical sensor is a well-placed solution, specifically for applications that have multiple points of discrete inspection but do not have tight part fixturing. This type of sensor is also good when different parts are run on the same line and require line configuration changes that would seriously hamper sensor arrays or require significant changeover or planned downtime to allow for changes in sensor placement. However, an optical sensor would not be as useful where a single discrete sensor or two could also solve the application. It would also not be as useful in applications that need complex inspection algorithms or complex internal logic. In these cases, a vision system would still be the better choice. Still, in most cases, the field is left wide open for optical sensors to be used during the manufacturing process to check for specification and quality adherence at each step of the production process. This also eliminates the need to install more complex vision systems for these types of applications to accommodate the needed functionality.

This new type of sensor provides aspects from both technologies to provide a number of benefits, including:
• Comparative simplicity, with simple configurations and interface but with multiple sensing functions within a single device.
• Faster set-up and operation, with the solid reliability of discrete sensing methodology.
• Lower overall cost to implement and maintain compared to vision systems at the high-end or multiple sensor arrays (including single purpose smart sensors) at the low-end.

Tom Draper is the manager, marketing programs, for Balluff Inc. Mark Sippel is the product marketing manager for Object ID with Balluff Inc.