Manual-to-automated vision saves costs

Labor-intensive applications in the industrial, medical, scientific, and military/aerospace markets are ideal candidates for automated machine-vision and/or vision-guided robotic systems.

Oct 1st, 2001
Th 74426
Click here to enlarge image

Labor-intensive applications in the industrial, medical, scientific, and military/aerospace markets are ideal candidates for automated machine-vision and/or vision-guided robotic systems. These systems not only eliminate the tedious, repetitive, error-prone, and costly tasks performed by operators, but also increase productivity, reliability, and precision. And, users are finding their investment in machine-vision-guided robotic systems is generally paid off in just one or two years.

In numerous applications, machine-vision and robotic-automation technologies are being merged for improved assembly applications. In this month's Product Focus, editor Andy Wilson explains how low-cost smart cameras, PC-based frame grabbers, and pattern-matching software are being integrated into vision-guided robotic systems (see p. 65).

In the automotive industry, many tasks involve the movement of large, heavy objects. To automate the production of Dodge pickup trucks, for example, DaimlerChrysler has developed an integrated robot and vision system that inspects racks of truck-bed parts, unloads different sized parts, and then correctly places the parts on a fixture for the next assembly step. This system, says editor Andy Wilson, not only adds flexibility, reliability, and accuracy, but also can adapt to changing parts geometries without the need for retooling (see p. 45).

The military has long sought a smart missile that could select one target out of several for destruction. In the past, error-prone self-guided missiles used laser light and a spatial light modulator as an optical correlator to compare all pixels in an input image to a stored template. Reports contributing editor Winn Hardin, university researchers in England have replaced the spatial light modulator with a DSP array that has simplified system design (see p. 51).

Manually mating a single-mode fiber to a laser diode is inefficient, expensive, and time-consuming. To overcome these shortcomings, discloses contributing editor Charles Masi, a systems integrator has developed a vision and motion-control system that uses a CCD camera for coarse alignment of the laser diode's beam and a linear variable displacement transducer for establishing a coordinate system relative to the laser diode's package (see p. 61).

When performing antiangiogenesis-drug testing in rat-retina specimens for cancer research, developers integrate microscopes, antiblurring algorithms, and automated micropositioning systems to build vision systems that rival more-expensive laser-scanning confocal microscopes. According to contributing editor Winn Hardin, these vision-control systems measure the microscopic changes in 50-µm-thick samples using 3-D volumetric imaging (see p. 29).

George Kotelly, Editor in Chief
georgek@pennwell.com

More in Robotics