System integrators are constantly searching for novel and improved ways of configuring systems using off-the-shelf and custom hardware and software. Sometimes, the design task is straightforward, such as using pixel-counting to read labels on bottles; sometimes, the requirements are complex, such as making several checks on car transfer cases or combining surface and volumetric renderings for 3-D modeling. Whether simple or difficult, all system designs mandate the use of proper lighting devices. This month's articles underscore these design concerns.
An industrial company that packages more than 400 types of bottles preferred an economical inspection approach to confirm that processed bottles had proper labels installed on both front and back surfaces. After a few system trials, it selected a pixel-counting-sensor approach that renders pass-fail judgments based on pre-established thresholds. As discussed by contributing editor Larry Curran, this inspection system is running smoothly 24 hours per day and nearly seven days a week, with as many as 12 product changes per day (see p. 22).
The cost of three-dimensional (3-D) modeling applications has steadily decreased due to continuing improvements in microprocessors, graphic boards, and open graphics standards. The results are improved 3-D algorithms and wider commercial graphics support on desktop platforms. In a medical-military collaboration, as reported by contributing editor Winn Hardin, surgeons at Massachusetts General Hospital in Boston have integrated Able Software 3D-Doctor tools and a custom PC-based platform to combine 3-D visualization with a mannequin's torso to replicate human internal organs (see p. 28).
To correct both product-assembly and inspection problems, a car power-transfer-case manufacturer searched for an automated machine-vision inspection system that could check four different cases per minute. The system had to verify the position and diameter of bolt holes, the presence of threads, and the measurement of flange dimensions. As described by Bob Emery and Ben Friend of AutomationSolutions, these requirements resulted in a complex vision platform that integrated a PC, a logic controller, special software, four cameras, and expansion capabilities (see p. 34).
According to nearly all system integrators, lighting is the most important parameter in developing a machine-vision system. Fortunately, numerous illumination technologies are available, such as fiberoptic, fluorescent, and light-emitting diode (LED). In this month's Product Focus, editor Andy Wilson explains why LED lighting devices are made the primary selection by system integrators (see p. 41).
George Kotelly, Editor in Chief[email protected]
