A Vision For Productivity

FAYETTEVILLE, Ark. - University of Arkansas industrial engineering professor Earnest Fant wants to help business owners, managers, and engineers decide when and how to implement optical technology, or "machine vision," and so he has written and released a new CD-ROM on the subject, Consider Machine Vision.

"Businesses are confronted with many challenges, such as worker shortages, worker retention, workplace safety, product quality, and productivity," Fant says. "Machine vision, which can range from the sophisticated devices that inspect products for surface flaws on an assembly line to a product scanner at the grocery store, can help solve these problems, increasing both productivity and profit."

The biggest problem presented by machine vision is how to decide if a particular optical product is suitable for the job at hand, Fant explains.

"Machine vision products can be bought in a variety of configurations, from low-resolution, low cost devices to highly sophisticated, but relatively expensive systems," he notes. "It is important to select the correct configuration for the job at hand, but business owners and managers frequently must rely on the seller for information to make a decision."

Fant hopes to make things easier for business owners, managers, and engineers by giving them some basic tools for making those decisions. In his CD-ROM, Fant explains the fundamentals that need to be considered in making an informed decision. Figures and more than 20 video clips are also included to illustrate concepts.

The idea for the CD-ROM grew from his frustration in looking for a textbook for his industrial engineering course, "Application of Machine Vision." The textbook he had used was more than 10 years old - hopelessly out of date in such a dynamic field - but no more current ones were available. He decided to publish as a CD-ROM because it would allow inclusion of video clips and be easier to keep current than a traditional paper book.

Standard "pick and place" robots are machines that have been used in U.S. industries for about 30 years. In 1995, there were approximately 20,000 robots in use in industry in the United States. By 1999, that number had risen to 100,000. They are generally used for applications that are demeaning, demanding, dangerous, or difficult for humans. For example, robots are used in feed mills to move 100-pound bags from the assembly line to pallets for shipping. Humans would quickly tire and could be injured by moving heavy loads and working in such a dusty environment.

By adding vision to these robots, their capabilities can be greatly expanded. These sensor-equipped machines can assemble and place components rapidly and with high accuracy -- up to 1/1000th of an inch. Additionally, because they do not tire, they do not make fatigue-induced errors. Currently machine vision is widely used, across many types of industries, from food processing to microchip manufacturing to automobile production.

Vision-equipped machines do not just replace human workers. They can also be used to enhance the working conditions for operators and improve their productivity. For example, adding an optical device over an assembly line that is linked to a screen beside the operator can allow the inspection of parts on the line at resolutions and angles that a human could not see unaided. This could allow an inspector to work more safely and efficiently, while increasing productivity.