Machine vision reads tea leaves

Mixed trends offer opportunities, and a court case clears the air.

Mixed trends offer opportunities, and a court case clears the air.

Signs of growth are promising for companies in imaging and machine vision. Critical industries that drive business—semiconductor, electronics, and automobile manufacturing, pharmaceutical production, and food and beverage processing—all look stronger in 2004. And sales of imaging-related equipment are rising, although profits are constrained.

These trends were among the many developments explored at the 12th Annual Business Conference of the Automated Imaging Association (AIA; Ann Arbor, MI), held Feb. 4–6 at the faux Italienne Portofino Bay Hotel in Orlando, FL. With about 160 attendees, the meeting is the year's best executive networking event for manufacturers of cameras, boards, and other imaging components, along with their distributors and system integrators.

During the past three years revenues for North American companies declined 15.1% while the number of machine-vision components sold increased 5.7%. These are some of the statistics used by Nello Zuech, president of Vision Systems International (Yardley, PA), to illustrate the complex and sometimes contradictory nature of the market. His annual survey of the North American machine-vision market will be completed this spring, but total revenues for 2003, including value added by system integrators, was $1.605 billion, and he says revenues should rise 8% a year to $2.44 billion in 2008.

Zuech says increasing numbers of North American companies think the majority of their sales will be in the Pacific Rim. However, capital spending for productivity does appear to be improving for consumable goods such as food, pharmaceuticals, and cosmetics, and these are unlikely to migrate offshore. In addition, the erosion in component prices seems to be lessening, but there is increasing resistance from end-users to pay for value-added integration services, especially as more end-users become familiar with vision technology. So the general feeling of most machine-vision companies seems to be that business conditions will be better in 2004, but still very demanding and competitive.

The crowd went wild

A rousing standing ovation is not what you might expect from a group of engineers-turned-executives, yet that's what victory in the Lemelson case produced at a lunchtime award ceremony. A special recognition award went to Cognex (Natick, MA), which along with Symbol Technologies (Holtsville, NY) and others had fought a long patent battle with the Lemelson Medical, Education & Research Foundation, Limited Partnership.

Although the inventor Jerome Lemelson died in 1997, the partnership has collected more than $1.5 billion in license fees from the machine-vision industry for patents covering machine-vision and barcode-reading technologies. On Jan. 23, Nevada U.S. District Court Judge Philip Pro ruled that the 14 patent claims are "invalid, unenforceable, and not infringed by Symbol or Cognex, or their customers."

Bill Silver, cofounder and chief technology officer at Cognex, accepted the award. He told me that the legal battle—which had been passed up by much larger companies—cost Cognex $5 million plus endless hours in Reno hotel rooms preparing the case. "Lemelson had a brilliant strategy of never making the cost of settlement [for licensing] more than companies were willing to pay," says Silver. "But they didn't expect to run into 'Dr. Bob' Shillman [chairman and CEO of Cognex], who felt it was just wrong, and was crazy enough to work 14 years on the case even if it didn't make business sense."

The judge, with no patent expertise, immersed himself in the technology before writing the decision. In dismissing the Lemelson patent claims, he wrote of one claim: "Additionally, the intrinsic evidence establishes that Lemelson's invention involves scanning by the use of a television or video camera and not scanning by means of a laser or CCD camera employed by Symbol or Cognex products. Indeed, the evidence clearly establishes that neither laser nor CCD cameras existed in 1956 or 1963, and that no one or ordinary skill in the art would have or could have described such scanners at that time." And that rebuke is just a taste of what brought the standing ovation. The industry owes Cognex and Symbol Technologies thanks for removing a burdensome expense and opening the door to further innovations.

CONARD HOLTON is editor in chief of Vision Systems Design; e-mail:

More in Detectors & Imaging