Application-specific optical fibers are designed to perform a particular function or set of functions within an optical system. The telecom boom affected the market for these fibers in such ways as advancing new applications, improving fiber manufacturing and reducing costs. Fibers that once were locked inside Bell Labs or that cost $50 or more per meter are now commercially available — and at much lower cost. Now a third generation of development is beginning that should advance optical power delivery and sensing as well as continue the optimization of communications fibers for a wide variety of applications.Optical fiber applications fall into three categories: data transmission, power delivery and sensing. Silicate fibers are available that transmit data and power at wavelengths from the deep-UV to 2.2 μm at very low power levels or at the limits of silica’s damage threshold.Fibers may be optimized to create optical circuits that branch, combine, attenuate and switch. With the development of rare-earth-doped fibers, for example, input signals can be optically amplified. The erbium-doped fiber amplifier revolutionized the design and economics of long-haul telecom systems, and other rare-earth-doped fibers are used in fiber lasers, which are expected to replace more than 60 percent of nondiode laser systems. Fiber Bragg gratings, once incorrectly predicted to become the technology of choice for dense wavelength division multiplexing filters, are enjoying new life in point and multipoint sensors and in smart structures. And nonlinear effects in fibers, such as Raman and Brillouin scattering, enable monitoring of temperature or vibration in a 20-km length of fiber.Although such applications had been proposed for years, what changed is that good systems engineering and the right fiber choices have reduced costs to a fraction of what they were prior to the telecom boom. This is because, first, the scale of global fiber production, which is tens of millions of kilometers per year, has driven costs down and production yields up. Second, new suppliers have generated intense competition and intellectual property. Third, better fiber-making processes, such as plasma chemical vapor deposition, are being applied to the development of specialty fiber designs.Figure 1. The plasma chemical vapor deposition process can produce complex, “hyperfine” refractive index profiles.That deposition process allows the fiber preform to be fabricated in very thin layers, with very precise control, which is useful in designs that require highly controlled, hyperfine refractive index and material profiles (Figure 1). Complex index and dopant profiles, comprising thousands of layers, can be rapidly locked into the glass at high yields and with high design fidelity. The process reduces manufacturing costs but, more importantly, enables rapid prototyping because it can often dial in the desired profile design on the first run instead of requiring multiple attempts.With novel manufacturing processes such as this, it is possible to ask a lot of a specialty fiber. For example, rare-earth-doped fibers also can maintain polarization and have a high-temperature coating, which is useful in military applications and for industrial processes that require organic buffer coatings that can withstand high heat. Depending on application needs, functions sometimes can be combined in a single fiber; for example, double-clad fibers can be photosensitive. But trying to meet too many needs can make the design impossibly expensive or just plain impossible.It is important to keep total system cost in mind from the beginning. Rather than thinking in terms of price per meter, remember that you often are buying a function — attenuation, amplification or dispersion compensation, for example — and not just a number of meters. Also, be sure to separate the cost of fiber development from the cost of the finished product. With a sensible design, a good preform process and statistical process control of the fiber draw, it is possible to manufacture the most exotic specialty fibers at costs low enough to launch many third-generation commercial applications.Meet the authorBill Beck of BTM Consulting in Hudson, N.Y., is director of North American sales for Yangtze Optical Fibre and Cable Co. Ltd. in China; e-mail: bill.beck@yofcna.com.