“They tell us, ‘We need a shorter commissioning time—the ability to have our line up-and-running as quickly as possible, with as little auxiliary support as possible,’ ” remarks Tom Kahn, product marketing manager with Schaumburg, Ill.-based vendor Omron Electronics LLC’s (www.omron247.com) Auto-ID and Vision Products Group.
Through connectivity to higher levels of control architecture within the plant environment, vision products have become more user-friendly. One recent solution Kahn mentions incorporates better intelligence and flexibility. “Tasks such as illumination and programming are simplified through the integrated HMI (human-machine interface) by real-time display to directly visualize lighting or programming.” Click-and-drag functionality with on-screen programming at the HMI terminal help make use easier.
To answer end-users’ demands, vision systems provider Cognex Corp. (www.cognex.com), of Natick, Mass., focuses on two areas, says Bryan Boatner, In-Sight product manager. One is reducing manual configuration. The other is improving the degree to which vision systems will communicate with factory-floor devices such as programmable logic controllers (PLCs), HMIs and robots.
Reducing configuration is largely an issue of software algorithms, Boatner explains. “If you’re a seasoned end-user of machine vision, or if you’re a beginner, there’s nothing more time consuming than having to tweak a number of parameters to get the software tools operating correctly,” he notes. Thus, he believes algorithms need to be designed to accommodate on-the-fly run-time changes.
Based on end-users’ feedback, Boatner’s company also preconfigures devices with several communications protocols. Those include add-on profiles, which enable PLCs to recognize vision sensors or systems without the need for manual configuration. “For example, with Rockwell Automation PLCs, we support add-on profiles. All you have to do is enter the
IP (Internet Protocol) address. Beyond that, it’s plug-and-play.”
Ben Dawson, director of strategic development for Billerica, Mass.-based Dalsa IPD (www.goipd.com), another vision vendor, believes that easy-to-use software is the main key for smart systems. But he considers three factors vital to make software easy to use.
“First, the user interface must be simple and intuitive,” Dawson says. To ensure this first target is met, he thinks three rules must be followed: “The user must not be left guessing as to what to do next, the software must present a simple ‘mental model’ to the user—that is, ‘If I do this, it does that, as expected’—and the software must not require a lot of memorization to use.” One way of accomplishing that, he notes, is enabling set-up and execution with buttons and graphics that have the “look and feel” of a good consumer product.
Make it smart
To account for the possibility that end-users might not have a machine-vision background, one solution is having built-in machine knowledge, Dawson remarks. “You might think of this as an “expert system,” but one limited to specific vision tasks.”
Finally, Dawson indicates that making systems easier to use requires expressing machine operations so a smart vision system “can do it in terms familiar to a technician or manufacturing engineer.” An example he notes is a caliper tool that conceptually works like a mechanical caliper or micrometer. That way, users who know traditional gauging and quality-control methods, but are unfamiliar with vision basics, such as edge detection, can more easily pick up on how to use the technology, he adds.
“Engineers or technicians can do most machine-vision applications with little background in vision,” Dawson concludes. But the key, he emphasizes, is designing the software for them.
C. Kenna Amos, firstname.lastname@example.org is an Automation World Contributing Editor.