The Model T, according to Henry Ford, was available "in any color you choose, so long as it's black." Introduced in 1908, the Model T revolutionized automobile manufacturing—Ford was not the first to build a car on an assembly line, but he perfected the system that enabled him to supply a motorcar for the masses (as long as they wanted black).
Today, nearly a century later, technology products and their manufacture have evolved to the point that almost any product can be customized in many different ways to address specific needs of an end user. Ironically, for many of "the masses," the choices now available have become almost overwhelming—a visit to your local cellphone store will make that point clear. But in the world of optoelectronics, the evolution of specialized technology has actually necessitated development of highly specialized manufacturing techniques to facilitate the practical implementation of many of the most significant ideas.
While semiconductor lasers have been around for more than 40 years, for example, reliability at higher powers in pumping applications at 806 nm can still be an issue. A novel approach to the processing of these devices during manufacture may improve reliability by stabilizing the laser-diode facets (see p. 69). And while CCD sensors are not new technology, customized CCD camera systems optimized for the relatively new field of proteomics offer a unique set of imaging features tailored to the protein separation technique, as well as producing digital image data that can easily be manipulated and stored on a computer (using customized software of course). The result is much more effective instrumentation in a very complex field (see p. 84). Machine vision is also not a new concept, although current implementations are often fairly basic, but development of fully autonomous robots will depend on yet-to-emerge highly sophisticated on-board optical sensing and processing. One aspect of building a better "eye" for robots involves implementing data compression in the physical layer of the sensor instead of after the image has been sensed, which is current practice (see p.77).
Ultimately though, success in producing customized photonic devices depends both on the materials themselves and the way in which they are handled. The feature articles in our Optoelectronics World supplement following page 92 highlight these basics, thus helping us all avoid the predicament of being able to make a laser any color you want as long as it's the 643-nm output of the original ruby laser.
