Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 235
Harnessing Light: Optical Science and Engineering for the 21st Century 6 Manufacturing Optical Components and Systems Introduction Since the early part of this century the manufacturing of optical components and systems has changed dramatically throughout the world, both in the types of products that are made and in the approach that is taken to making them. Once devoted entirely to passive image-forming components (such as lenses and mirrors) and to the instruments made from them, the industry now also manufactures a wide range of active elements such as lasers and optical sensors. Until recently, the industry depended heavily on a craftsman-style approach to manufacturing, with much of the work being carried out on an order-by-order basis by very small businesses. As new mass consumer markets have emerged that rely on optical technology—such as compact disk (CD) players and laptop computer displays—the implementation of high-volume mass-manufacturing techniques similar to those of the electronics industry has revolutionized this segment of the optics industry. To take just one example of this new manufacturing technology, more than 100 million diode lasers are now produced each year, on highly automated production lines. The availability of these inexpensive diode lasers has revolutionized entertainment (in CD players), made high-quality printing affordable for small businesses and home users (in laser printers), and made possible numerous other new products that together account for hundreds of billions of dollars in global business revenue each year. These changes in manufacturing are exciting, but they are reflected most prominently in the globalization of the optics industry, rather than in the domestic development of U.S. industry. Indeed, almost all mass
OCR for page 236
Harnessing Light: Optical Science and Engineering for the 21st Century production of optical components and systems now takes place outside the United States. There are only a handful of large U.S. optics companies engaged in the volume production of optical components, most of them in the plastic lens component business. This U.S. trend toward specialty products and small companies has been due in large part to the special technical needs of the Department of Defense, which has long been a vital customer for the industry. Government programs such as Small Business Innovative Research (SBIR) have also encouraged the formation of small, innovative optics companies. The main strength of the U.S. optics industry is now in high-precision manufacturing of low-volume specialty optical components and devices with high added value. This strategy has produced a strong industry based on the diverse activities of many small companies but lacking the manufacturing base required for expansion into mass consumer markets. There are several thousand small optics and optics-related companies in the United States, with an average of 50 or 60 employees each. Together they account for more than 200,000 jobs and annual net revenues of about $30 billion for optical components and systems (excluding ophthalmics).1 Yet even these impressive statistics do not adequately indicate the strength that small businesses provide to the U.S. optics industry as a whole, by making available a broad range of technical skills. A key finding of this report is that despite the optics industry's significant contribution to the U.S. economy, this contribution comes in so many small pieces that it is not usually fully recognized and understood. The enabling character of optics, a repeated theme of this report, is an especially important consideration for the manufacture of optical components. The value of a component such as a laser diode or an aspheric lens is usually small compared with the value of the optical system that it enables. It is even smaller compared with the value of the resulting high-level application. Advances in the manufacturing of optical components are greatly magnified into improved capabilities and economic advantages at the systems and applications level. Advanced optical components cannot be considered commodity items. This chapter addresses two distinct challenges. First, how can we maintain and strengthen the U.S. optics industry's leadership in high-precision manufacturing of low-volume specialty products? Second, how can we ensure the U.S. optics industry's ability to compete internationally in the increasingly important mass markets, especially the new mass markets that continue to emerge? Following a brief history of optics manufacturing in the United States and a short overview of the current state of the industry, the chapter divides into two main parts: (1) low-volume manufacturing of high-performance specialty products 1 These numbers are based on a sample of the companies listed in the annual Photonics Directory.
OCR for page 237
Harnessing Light: Optical Science and Engineering for the 21st Century and (2) high-volume manufacturing for the mass markets. The chapter ends with a discussion of some crosscutting issues, such as metrology and design, and the industry's composition, size, and growth. A Brief History Before about 1910, the U.S. optics industry consisted of just a few manufacturers of optical instruments such as binoculars and inspection equipment. Virtually all such products were imported from Europe. World War I stimulated demands for a domestic capability, and the need to provide components for these instruments was the basis for the U.S. optics industry's growth throughout the early part of the century. The 1920s and 1930s supported several medium-to-large optical companies, such as Bausch and Lomb, American Optical, and Eastman Kodak—high-volume producers of both traditional and new optical instruments. A well-organized photographic industry provided almost all the cameras demanded by U.S. consumers. Most microscopes, binoculars, telescopes, and optical inspection equipment were also manufactured domestically. The needs of the military during World War II placed significant demands on the industry's capabilities, and when military contracts ceased abruptly at the end of the war, most optics companies fell on hard times. Demand for cameras and other optical instruments for consumer and civilian uses grew, but Japanese and European competitors could satisfy this demand more cheaply than most U.S. companies, many of which succumbed to the competition. The remaining domestic camera and instrument manufacturers cut costs by turning to component suppliers in the Pacific Rim, first in Japan and more recently in China and Malaysia. From the 1950s through the 1970s, the industry became increasingly divided, with overseas suppliers dominant in the high-volume markets and U.S. industry focused on assembly, systems building, and low-volume specialty components. Small companies came to dominate the U.S. optics industry. In 1960, the invention of the laser spawned an entirely new segment of optics manufacturing, a segment that has grown astonishingly. Technologies developed to take advantage of the laser's capabilities have led to additional major markets for optical fibers, optical communications systems, optical sensors, and a broad range of other new applications. Mass U.S. markets for these applications have been based on aggressive growth in overseas manufacturing of the basic components. An Overview of the Industry Today The nature of the optics industry continues to change. Mass production techniques are used to manufacture components for an increasing
OCR for page 238
Harnessing Light: Optical Science and Engineering for the 21st Century number of high-volume optics applications. Among the products manufactured in this way are optical fiber for telecommunications and flat-panel displays for computers. Most of this type of manufacturing currently takes place overseas, not in the United States. At the same time, demand remains strong for high-performance specialty products that are manufactured in small numbers. There are three main markets for these items: (1) the military, (2) other high-technology scientific and government programs, and (3) specialized industrial applications. Many high-performance military optical systems have very specialized capabilities but low production volumes. Some federal facilities for civilian research and development have similarly specialized needs. A key private-sector market for high-precision optical systems is the electronics industry, in which a relatively small market for photolithography systems enables the huge semiconductor business. The United States excels in this high-value, low-volume portion of the optics industry. Most of the industry that serves the low-volume, high-accuracy component market remains dependent on very traditional fabrication methods, although it is increasingly facilitated by high-quality interferometric test equipment. This sector of the industry, made up mostly of small companies, faces increasing competition and must adapt to the new global marketplace. To maintain market share as overseas competitors improve their accuracy, domestic manufacturers will have to develop and use more deterministic fabrication methods that achieve the same results at lower cost with fewer high-skill workers. For each of these types of manufacturer, an important element in the future growth of the industry is the growing integration of passive image-forming components with active sensors and light processors. The acceleration of this trend will mean a corresponding integration of the traditional optical component industry with the developers and suppliers of electrooptical materials and devices. The challenges of the future will require new, faster, more flexible approaches to optical component fabrication, with less reliance on skill-intensive, iterative production methods. Some programs have already been established to promote this goal. For example, the Center for Optical Manufacturing has developed a series of computer-controlled generating machines that use diamond tools to produce accurate surfaces on glass elements. The Manufacturing Operations Development and Integration Laboratory (MODIL) has developed techniques for fabricating certain specialized laser components. Similar approaches are being implemented overseas. It is not clear, however, that such methods will be enough to revitalize U.S. production of high-volume general-purpose optical components, because most of the small shops that currently dominate the U.S. optics industry lack access to the investment capital necessary to upgrade their equipment.
OCR for page 239
Harnessing Light: Optical Science and Engineering for the 21st Century Collaborative programs in optics manufacturing should include universities so that students are trained in the latest technical solutions to production problems. The Department of Defense (DOD), the National Institute of Standards and Technology (NIST), and the Department of Energy (DOE) national laboratories should establish together a cooperative program that provides incentives and opportunities to develop new ideas into functioning methods for optics fabrication. A critically important asset of the U.S. optics industry remains its strength in optical design. U.S. software companies set the world standard for optical design programs, although their products are of course widely used overseas. The development of sophisticated lens design programs, with good interaction with the designer, is remarkable. Programs that will run on a high-level personal computer now give any optical engineer access to modern design tools, and this easy availability has stimulated a widespread interest in optical design. There is as yet little integration of active components into the design process, however, and comprehensive software for physical optical design is still at a relatively rudimentary stage. U.S. accomplishments in optomechanical computer-assisted design (CAD) and thermal analysis software would be even more effective if fully integrated into optical design software. Manufacturing of optical components and systems requires a large skilled and semiskilled workforce, and emerging new mass markets will increase the optics industry's need for trained workers. The quality and availability of optics training at the technician level is a widespread concern. A key challenge for the future is the establishment of standards for the interchangeability of optical components, which is an important driver for cost-effective manufacturing. U.S. participation in international standards-setting activities lags far behind the activities of foreign organizations. Low-Volume Manufacturing of Specialty Optics There continues to be strong demand for high-performance specialty products that are manufactured in small numbers. For many of these products, the customer is the government, especially DOD, but certain high-value items are also vitally important in the commercial sector. Specialized high-value applications, such as lenses for photolithography, continue to be an area in which the U.S. optics industry can excel. As described in Chapter 4, military optical systems tend to have high-performance and specialized requirements but low production volumes (Joint Precision Optics Technical Group, 1987). For example,
OCR for page 240
Harnessing Light: Optical Science and Engineering for the 21st Century forward-looking infrared (FLIR) systems require expensive infrared-transmitting materials as well as environmentally resistant surfaces, coatings, and mountings. Ring laser gyroscopes require low-scatter surfaces and very high-precision optical components. High-performance aircraft and missiles require unusual aspheric components, conformal to the shape of the airflow. Affordability is becoming increasingly important to the Department of Defense, but despite its wish to use commercial products off the shelf where possible, DOD supports design and manufacturing development for a number of specialized optical technologies. The volume of demand for such items, even including the commercial applications, is often too small to ensure the necessary development of fabrication techniques by industry alone. DOD should continue to maintain technology assets and critical skills in optics manufacturing in order to meet future needs. Some government projects require so many specialized optical components that they have a significant impact on the entire optics industry, despite the low volume for each of their individual components. Among these are the National Ignition Facility and the Atomic Vapor Laser Isotope Separation (AVLIS) program. These two DOE programs will consume thousands of medium-to-large optical components with high-precision surfaces and coatings resistant to high-power lasers. The overall size of these programs allows the private sector to plan some investments in improved machinery and processes. Photolithography for manufacturing electronics is a key private-sector use of high-precision optical systems. The production of short-wavelength photolithography systems of ever-higher quality is essential for continued growth of the semiconductor industry. The Moore's law trend of increasing semiconductor miniaturization will drive photolithography through deep ultraviolet (UV) wavelengths and into the soft x-ray region by the turn of the century. At present, most imaging tools are produced overseas, but there are opportunities for U.S. industry to take the lead as systems move into the far UV, if economical methods can be found for producing moderate-sized aspheric surfaces with an accuracy better than 1 nm. Specialized applications such as these incorporate a wide variety of traditional and modern optical technologies, each with its own manufacturing issues. Spherical Lenses The curved surfaces of a lens cause rays of light from a point on a distant object to come to a focus. A single lens with spherical surfaces, although quite economical to manufacture, forms an image that is not a perfect point (see Figure 6.1). Optical design has traditionally been a search for combinations of spherical-surfaced components, made of
OCR for page 241
Harnessing Light: Optical Science and Engineering for the 21st Century FIGURE 6.1 A single spherical-surfaced lens (left) forms an image that suffers from spherical aberration. To reduce this effect, a typical photographic or video lens (right) consists of many elements. different types of optical glass, that result in nearly aberration-free images. In general, the wider the field of view or the more extended the spectral range required, the more elements will be needed. The traditional approach to making spherical surfaces has been surface lapping, which can produce high-quality polished surfaces that deviate from the designer's specifications by as little as a few hundredths of a wavelength. This lapping or averaging method has been very successful in fabricating spherical and flat components, but it is by nature a time-consuming and craftsman-intensive activity. Improvements currently being investigated are directed toward deterministic fabrication, in which the accuracy of surface production is inherent in the machine carrying out the process rather than in the time-varying lapping of surfaces. Processes that are successful in finishing unusual materials, including optically active materials, have become more important. There have been several attempts to improve and modernize the methods used for serial production. These approaches, however, such as high-speed surfacing, molding, and automated test and assembly machines, are usually directed at reducing the cost of a specific product. The improved production capability rarely extends to other products. A major improvement has been the very inexpensive production of plastic components. The use of plastic components is currently limited to systems of moderate and low quality, however, and to a limited range of environmental conditions. At least in the near term, most optical systems will continue to require glass components. Currently, only one major domestic research and development program is directed toward the versatile production of economic spherical components. This is the Center for Optical Manufacturing (COM) at the University of Rochester, which has made significant progress in the development of high-speed machines to generate surfaces that require
OCR for page 242
Harnessing Light: Optical Science and Engineering for the 21st Century minimal polishing. COM has also developed a promising experimental polishing process called magnetorheological finishing. In magnetorheological finishing, the conventional rigid polishing lap is replaced by a suspension of magnetic particles and polishing abrasives. A magnetic field locally stiffens the fluid, creating a polishing spot that can be scanned over the part to polish it and correct its surface figure. Although most spherical optical components are produced for use in imaging systems, other applications are also important for a healthy U.S. optics industry. Many specialized components are needed for laser systems, data storage systems, telecommunications equipment, a variety of analytical instruments, and endoscopes or other optical devices for minimally invasive surgery. The U.S. catalog houses that distribute stock produced overseas meet only part of this need. Aspheres If the surfaces of an image-forming component are allowed to be nonspherical, a major advance occurs. The addition of definable high-order curvature to the usual second-order spherical surface permits independent correction or balancing of spherical aberration. This leads to a reduction in the number of lens surfaces needed for aberration-corrected imagery and permits simpler, better optical systems (see Figure 6.2). Asymmetrical aspheres are also becoming important, especially in conformal applications, in which the outer surface of an optical component must conform to the aerodynamic shape of an aircraft or missile. Among the diverse applications of high-precision aspheres are military aerospace systems, optical data storage, photolithography, and astronomy. Lower-precision aspheres have an even wider range of application, including photography and video imaging (especially zoom lenses), projection television, medical instruments such as endoscopes, telecommunications, and document scanners and printers. At the low-end of the market, aspheres find use in such applications as condenser elements for illumination. Some of these areas of application present valuable economic opportunities for the U.S. optics industry. The trend toward more compact optical systems with active or movable components—zoom lenses, for example—is increasingly driving designers toward aspheres. The future goals of the semiconductor industry (which will require asymmetrical reflective components for deep-UV photolithography), as well as other industrial and defense applications, cannot be achieved without the ability to produce high-quality aspheres cost-effectively. FIGURE 6.2 A single element with an aspheric surface can have significantly reduced aberration (compare with Figure 6.1).
OCR for page 243
Harnessing Light: Optical Science and Engineering for the 21st Century Aspherical surfaces cannot be produced by the traditional methods used for manufacturing spherical-surfaced components. Three technologies are required to manufacture precision aspheres: machining, polishing, and metrology. Successfully generating, polishing, and testing aspheres all require considerable skill, and investment in all three is mandatory for successful production. The two current options for machining defined aspheric surfaces are single-point diamond turning and computer-controlled (CNC) generation. Diamond turning is used to make aspheres in ductile materials such as metals, but it leaves an intrinsic surface roughness that must be removed by polishing if the component is to be used at visible or near-infrared wavelengths. CNC generators are used to grind brittle materials such as glass. Several countries have programs directed toward defined-shape generation of aspheric surfaces. These programs are beginning to produce numerically controlled machinery that makes the use of multiple aspheric surfaces feasible. In the United States, a major new program funded by the Defense Advanced Research Programs Agency (DARPA) is investigating the fabrication of conformal optical components. Except in some infrared applications, machined aspheric surfaces usually require a computer-controlled polishing or finishing step. This step has two purposes: to correct the surface figure to the proper shape and to smooth its microroughness to reduce light scattering. The achievable accuracy of aspheric surface figuring has improved a hundredfold in less than a decade, from 0.5 μm [root mean square (rms)] in 1988 to about 5 nm (rms) in 1996. A variety of approaches have been developed based on loose abrasive polishing. These include passive and active flexible laps, deformable rigid laps, and small tool polishers. All of these approaches aim to solve the problem that a rigid lap, as is used in the conventional polishing of spherical elements, will not maintain good contact with an aspheric surface as it moves over the part. Magnetorheological finishing is also under development, as mentioned above. Ion figuring has proved to be a predictable method for final shape correction, but it does not reduce surface roughness. The third requirement is metrology of the surface figure. Comparing surface measurements to the design of the component produces the data needed to drive computer-controlled polishers. Thus, the metrology system not only qualifies the finished part but also is essential during the polishing process. Contact and noncontact profilometers are used, but these systems are slow, and more importantly, they measure the surface only along widely separated one-dimensional traces. An interferometer coupled with a null optic (refractive, reflective, or diffractive) is preferable because it provides an accurate full-area test of the surface profile. The null lens must be calibrated before
OCR for page 244
Harnessing Light: Optical Science and Engineering for the 21st Century use. Asphere manufacturers find the use of null optics essential, but it is common for them to lament the high cost and long lead times. All three of the above technologies need improvement, but the technology deficit is most severe in metrology, followed by polishing, and least severe in machining or generating. Also important for the future fabrication of precision aspheres will be the trend away from rotationally symmetric aspheres or their off-axis sections and toward generalized aspheres with little or no symmetry. It is vital to the future of the U.S. optics industry that domestic production of aspherical optical components be made more cost-effective. Moderate-quality aspheres typically cost four to five times more than comparable spherical elements, so simply eliminating a spherical component from a design is rarely enough to make the economic case for using an asphere. Performance, size, or weight must be an additional driver. Thus, asphere fabrication technology has until now been driven by the needs of specific applications. More widespread application will require continued moves toward technology-based and computer-controlled manufacturing processes. These capabilities tend to lie in larger companies, and as a result, only a few U.S. manufacturers are currently able to produce aspheres economically. The ability to produce aspheres at a cost less than twice that of a spherical surface would open wider markets for such components. Without an incentive to develop a domestic capability, there is a real risk that precision asphere manufacturing will move offshore. Government agencies should continue to support the activities necessary to introduce cost-effective precision aspheric components into both military and commercial products. Computer-Controlled Deterministic Grinding and Polishing The primary reason high-volume manufacturing of image-forming components has left the United States is that current manufacturing techniques for these products are still operator resident, art driven, and based on labor-intensive machinery that often dates from the 1940s. As a result, the quality of the output depends on the skill and experience of the operator, rather than on computer controls and a scientific understanding of the manufacturing processes. Revitalizing the industry will require a move toward transferable manufacturing processes that are based on smart operators and computer controls. This technology will be more capital intensive and less operator dependent than the current approach. Computer-controlled, deterministic processes will also lead to better consistency—all parts of identical quality and the first part as easy to produce as the last—and to significantly faster cycle times. A model to emulate is the metalworking
OCR for page 245
Harnessing Light: Optical Science and Engineering for the 21st Century industry, in which computer-controlled machinery is now the norm and handbooks are widely available that list feeds and speeds for various materials and configurations. To begin addressing these needs, three universities and the American Precision Optics Manufacturers Association, with Department of Defense sponsorship, have jointly established the Center for Optics Manufacturing. COM's major thrusts are the integration of computers into optical component manufacturing, the development of computer-controlled optical component manufacturing equipment, the development of deterministic processes using this equipment, and the transfer of the resulting technology to industry. Some of the key challenges being addressed include improved inspection techniques for in-line process control, new machine geometries for making aspherical parts, and improved tools for optomechanical design. This new deterministic approach creates some educational challenges. Advanced-degree programs must integrate mechanics and materials science into the traditional optics manufacturing curriculum. At the technician level, mechanical and computer training are required as well as training in optics, to ensure that machine operators' skills are appropriate for computer-controlled fabrication equipment. Diffractive Elements The use of diffractive optical elements, in which light is manipulated by a microscopic pattern on a surface rather than by its macroscopic curvature, is a fundamentally new approach with tremendous potential. Diffractive components exploit the wave nature of light to form images by the effect of a series of zones generated on a base surface. Some of the features of such components are their low weight even at large apertures, their ability to correct spherical and other aberrations, and their reduced need for exotic materials. Applications generally use diffractive elements in combination with refractive and reflective elements. The fundamental building blocks for diffractive technology are in place, and as a result, interest in this approach has surged recently. Designs that combine aspheric lenses and mirrors with surface-relief diffractive lenses offer significant potential to improve performance and reduce size, weight, and cost. Among their applications are head-mounted displays, advanced sensor systems, laser and broadband imaging, optical interconnections for high-speed data transfer, optical data storage, optical correlators for target acquisition systems, consumer digital imaging applications, and precision testing of complicated optical systems. Diffractive optical elements have significant market potential. Estimated at between $15 million and $20 million in 1995, the market is expected to grow to between $150 million and $200 million by the turn of the century. Almost one-quarter of the market will be in
OCR for page 264
Harnessing Light: Optical Science and Engineering for the 21st Century evaluated for tolerances, so they do not work when assembled or are impossible to build. There are less than a thousand expert optical design engineers in the United States, but many more people have access to design software. Experience plays a significant role in this field; it takes about 3 to 5 years of full-time effort for a competent designer to become comfortable with a wide range of applications. Third, optical systems can now be fully analyzed for very little cost. The number of rays a designer can afford to trace no longer limits his or her abilities. The total design process begins with a meeting of the designer and the customer to define the goals of the system, including such factors as performance, cost, package, and delivery. About 25% of the design effort is usually spent in communication with the customer to ensure that the product is feasible, that the estimated cost is compatible with the budget, and that the design is indeed what the customer wants. Examination of manufacturability should start in midproject. This includes tolerancing limits of the design and checking compatibility with mechanical constraints, mounting, and environmental issues such as temperature range. The design is then fine-tuned by taking into account all known issues. The total design time can range from a few weeks to 6 months depending on a number of factors including the complexity of the system and the cost sensitivity of the manufactured product. Surprisingly, despite the availability of tolerancing and manufacturability tools in design packages, many designers fail to examine these issues, concentrating only on analyzing and optimizing the design's image and aberration. In most cases, design software can model, specify, and predict the performance of complex surfaces that are beyond what can be accurately fabricated, tested, or aligned. Designers must be aware of these limitations. Better tolerancing is a challenge for the optical design community. Areas that need improvement include communication of the fabrication and testing limitations of the optics shop, communication of the tolerancing set and compensators, better understanding of the manufacturing cost breakpoints, and more enthusiasm for the tolerancing process. New tools for computer simulation of multiple-step assemblies will result in better correlation between computer-predicted and as-built performance. The fourth major ramification of the revolution in design software is the possibility of global optimization, a major breakthrough that allows multiple local aberration minima to be traversed in a single computer run. Historically, local minima have been a common problem facing optical designers. The system they have designed may be the best of all similar systems for the given application, but there is no guarantee that other classes of design (using more or fewer elements, other types of glass, and so on) would not produce superior results. Real problems with many variables and many constraints-such as designing a lens
OCR for page 265
Harnessing Light: Optical Science and Engineering for the 21st Century that fits within a volume or length constraint—can now be analyzed overnight on a desktop computer that generates a family of alternate solutions of different types. Global optimization is one of the most significant innovations in optical design in the last decade. Despite these changes, no foreseeable advances in design software will eliminate the need for human intervention. Incremental designs, minor improvements, or small configurational changes in existing systems can result from optimization techniques built into a software package. Nonincremental designs represent entirely new configurations of optical elements and cannot be obtained automatically. Truly innovative nonincremental designs represent new technology and require invention by an expert optical designer. By one estimate, only about 3% of design jobs result in nonincremental improvements. New optical technologies continue to pose new challenges for the optical design community. Optical systems now range from surgical endoscopes a few millimeters in diameter to space-based telescopes 25 m wide. Systems containing adaptive elements force new optimization guidelines on the designer. Improvements in software are necessary for the design and optimization of nonimaging components. Optical design tools must advance to handle integrated optical and mechanical design; for example, standards for representing precision surfaces are needed that are compatible with mechanical design programs. In the future, as computation becomes ever cheaper, end-use image simulation will be employed instead of ray fans for system analysis and optimization. Resources must be found to incorporate all of these changes into optical design software packages. The optical design community is small, and optical design is taught formally at only a few universities, which has limited the number of people dedicated to developing design software. Rarely are students given more than introductory training in optomechanical design. The backlog of potential productivity enhancements from improved designs is measured in years. Role of Metrology Advances in optical metrology over the past decade have opened up vast new areas in optical fabrication. There is an old saying that "you can't make it if you can't measure it." A corollary to this has now been well established: "If you can measure it, you can indeed make it." Advances in fabrication technology have sometimes outstripped our measurement capabilities. A lack of routine, cost-effective, timely metrology solutions is often the bottleneck in the manufacturing process. Much of the advance in metrology is due to advanced light sources, improved sensors, and the tremendous increase in computing power and reduction in computing cost. The availability of a variety of light sources at useful wavelengths, the development of two-dimensional
OCR for page 266
Harnessing Light: Optical Science and Engineering for the 21st Century high-density detector arrays, and the computational and storage capability of today's machines have made practical advances that previously were possible only in theory. Continued progress will lead the way in the development of new metrology systems. Its strength in computing puts the United States in a very strong position internationally for optical metrology. Previously unprecedented surface finishes have been made routine by commercially available noncontact surface profilers and scanning-probe microscopes that measure to the angstrom (Å) level. Even for large-scale optical components, such as mirrors for the Advanced X-Ray Astrophysics Facility (AXAF), x-ray test data have established that the fabrication metrology has an accuracy of better than 250 Å over a 1-m scan. Future developments in optical metrology must embrace conflicting requirements. There is an ever-increasing desire for higher spatial resolution at the same time as larger and more complete area coverage. The ability to manufacture diffractive optical components with high throughput requires the ability to sculpt profiles on submicron structures to accuracies on the order of tens of angstroms. The advancement of microlithography requires concomitant advances in metrology. Even on large scales, optical components such as those required for space-based observatories are affected by angstrom-level height variations over spatial scales in the hundreds of microns. The need to maintain special environmental conditions for high-precision measurements is a huge driver of system costs, and the final measurement accuracy is often limited by how well these environmental conditions are maintained. Consequently, the development of a metrology system insensitive to environmental conditions would be of tremendous benefit. Some other areas for future technology development include these: Rapid and inexpensive measurement of aspheres and anamorphic elements; Nondestructive evaluation of subsurface damage; Measurements of material homogeneity on large blanks in an unfinished or raw state; In-process metrology for material removal rates, dimensional changes of the blank, surface figure, subsurface damage, and surface finish; Metrology for surface feature generation in microlithography; In-process metrology for the deposition of well-defined multilayer coatings over large areas; Figure measurement for meter-sized parts with absolute accuracy on a scale of tens of angstroms; Metrology for diffractive optics;
OCR for page 267
Harnessing Light: Optical Science and Engineering for the 21st Century Rapid remote and direct measurement and quantification of both particulate and molecular contamination; and Development of sophisticated computer-generated null lenses. The current level of R&D support in the United States in these areas, whether by industry or government, is a strong ground for concern about continued U.S. technological leadership. There are no recognized calibration standards for surface roughness, scattering, or cosmetic defects. In addition there are no standard techniques for certification or calibration of holograms or binary optics to be used for optical testing. The quantification of surface defects and imperfections that affect system performance (scratches and digs) should be redefined along with other areas in an international standard. It is necessary to adopt such international standards in order to remain competitive globally. The increasing sophistication of metrology has made it difficult for many engineers and technicians to understand the meaning of data provided by current instruments. Results are often misinterpreted, and this problem will be compounded in the future as even more complex instruments are developed and the amount of data processing required makes the results still more remote from most users' intuition. Therefore, it is necessary to develop and incorporate practice-oriented instruction aimed at creating an in-depth understanding of state-of-the-art metrology as it relates to what is being measured and to the limitations of its use. Also, because software algorithms, which are not accessible to the user, are so important in advanced measurement instrumentation, instrumentation manufacturers must continue to provide instrument-specific training and user-friendly software. Metrology training should be done at the undergraduate and the graduate levels. Similarly, technician-oriented courses should be developed for technical schools and two-year degree programs. Standards The issue of standards has been mentioned several times in this chapter as a key challenge for the future. For example, standards for physical design (e.g., support, stress, and ease of assembly) are an issue in the packaging of photonic components. Photonic integration will require well-defined standards for form, fit, and function. Integrating optical design with mechanical design will require standards for the representation of precision surfaces. In the area of metrology, there are no recognized calibration standards for surface roughness, scattering, or cosmetic defects. There are no standard techniques for certification or calibration of holograms or binary optics to be used for optical testing. There is a need for redefined standards for the quantification of scratches and digs. Standards for
OCR for page 268
Harnessing Light: Optical Science and Engineering for the 21st Century the interchangeability of optical components are an important driver of cost-effective manufacturing. Successful competition in international markets depends on establishing workable international standards in areas such as these. Unfortunately, U.S. participation in international standards-setting activities lags far behind the activities of foreign organizations. Virtually all new optics standards have hitherto been developed overseas, with no support from U.S. industry or the U.S. government. As a result, the U.S. optics industry has been a follower, not a leader, in adapting to new international opportunities. Active government and industrial participation in the setting of strong international standards for optical components is especially important because of the diversity of the U.S. optics industry. Government agencies and the optics community should recognize the importance of optics standards, especially their significance in international trade. The U.S. government should participate actively in the setting of such standards. NIST should be given the funding necessary to take the lead in this area. Size and Composition of the Optics Industry There is no satisfactory comprehensive source of data on the optics industry. No single professional or trade organization represents the industry as a whole, and the industrial data collected by government agencies are of limited use because their classification scheme does not clearly identify optical products. Nevertheless, certain facts are clear. As mentioned in the introduction to this chapter, large companies do not dominate the U.S. optics industry. Some large companies, such as AT&T, focus on integrating optics at the systems level, whereas others, such as Kodak, focus on systems and components. Small entrepreneurial companies play a critical role in developing the component technology (Council on Competitiveness, 1996). Optoelectronics appears to be the most rapidly growing segment of the optics industry. The most comprehensive source of recent data on optoelectronics manufacturers is a survey of 106 U.S. companies conducted between July 1992 and February 1993 by the Bureau of Export Administration (BEA, an agency of the Department of Commerce). Of the 106 companies in the BEA survey, 77 were primarily manufacturers and 17 performed both R&D and manufacturing. (To protect proprietary information, BEA did not identify individual companies.) For the purposes of the survey, BEA defined optoelectronics as all systems, equipment, and components that emit, modulate, transmit, and/or sense light or are dependent on the combination of optical and electronic devices (see Table 6.1) (U.S. BEA, 1994). According to the survey, in 1989 these 106 companies had net sales of $4.4 billion. Respondents projected that by 1995 this figure would
OCR for page 269
Harnessing Light: Optical Science and Engineering for the 21st Century TABLE 6.1 Composition of the Optoelectronics Industry Category Products Fiber-optic communications Transmission, amplifiers, cable television distribution, optical modulators, switches, fibers, multiplexers, connectors, transmit and receive modules Fiber-optic information equipment Optical processing units, memory or storage devices, bar-code readers, printers, image processing, interconnects, faxes, displays Industrial/medical equipment Machine vision, optical test and measurement, night vision and surveillance, laser processing equipment, nonlaser medical equipment, lasers Nonmilitary transportation equipment Automotive interior displays, traffic control systems, fly-by-light, cockpit displays, lidar for sensing turbulence, optical gyroscopes Military equipment Fiber-optic ground and satellite communications, lidar, optical gyroscopes, FLIR, night vision, munitions guidance, laser weapons Consumer equipment Televisions, video cameras, CD players, home faxes, appliance displays Subsystems/components Photo detectors, semiconductor light sources, hybrid optical devices TABLE 6.2 Sales of 106 Surveyed Companies in Optoelectronics Sector of Optics Industry 1989 1990 1991 1992 1993 E 1994 E 1995 E Net sales ($ million) 4,412 4,974 5,576 5,873 6,430 7,134 7,921 Change from previous year (%) — 12.8 12.1 5.3 9.5 10.9 11.0 NOTE: E = estimated. Source: U.S. Bureau of Export Administration (1994), Table IV-4. approach $8 billion, which represents an average annual increase of 10% (see Table 6.2). This growth rate is probably representative of the optoelectronics industry as a whole, but the survey was not all-inclusive, and so the exact size and composition of the industry are not known (U.S. BEA, 1994, p. iii-2). Another source of data is the annual survey of manufacturers (ASM) conducted by the Bureau of the Census. This survey is one of the primary sources of information regarding the performance of the U.S. manufacturing industry. The most recent ASM, based on a sampling of approximately 58,000 manufacturing establishments, estimates the value of 1995 shipments for approximately 1,750 classes of manufactured products (U.S. Bureau of the Census, 1997). Two of these product classes are fiber optic cable and laser systems and equipment; each shows an average annual growth rate of about 15% in recent years (see Figure 6.4). Recent information in the trade press indicates that worldwide 1996 sales of industrial laser systems were $1.5 billion, a 29% increase from
OCR for page 270
Harnessing Light: Optical Science and Engineering for the 21st Century FIGURE 6.4 Shipments of fiber-optic cable and laser systems and equipment by U.S. manufacturers. (Source: U.S. Bureau of the Census, 1997.) 1995. A further 23% increase is expected in 1997. U.S manufacturers have a 35% share of this market by revenue—about $400 million in 1995 (Belforte, 1997). These growth rates are about twice those derived from the 1995 ASM, whereas the U.S. market size estimate is substantially smaller than the Census Bureau's estimate. The Optoelectronics Industry Development Association (OIDA) estimates that in 1995, world production of optoelectronics was $64 billion, of which U.S. production was $12.5 billion (see Table 6.3). OIDA predicts that the optoelectronics market will continue to grow by about 10% per year, resulting in a world market in 2013 that approaches $500 billion. TABLE 6.3 Optoelectronics Market Segments and Their 1995 Production ($ billion) World U.S. Cathode-ray tube displays 16.3 1.8 Data storage (media only) 13.0 1.0 Flat-panel displays 11.5 0.2 Sensors (including imaging) 7.0 1.0 Lightwave transmission, interconnects, etc. 6.0 3.5 Fiber, passive 6.0 2.0 Military 2.5 2.0 Lighting 1.5 1.0 Solar cells 0.1 0.0 Total 63.9 12.5 Source: A. Bergh, Optoelectronics Industry Development Association.
OCR for page 271
Harnessing Light: Optical Science and Engineering for the 21st Century BOX 6.3 INDUSTRIAL CLASSIFICATION CODES The basis for collecting industrial statistics such as those described in this section is a system of product classes or groupings of the individual products of an industry. Since 1939, product classes have been designated by a system known as the Standard Industrial Classification (SIC), which was established by the Department of Commerce in concert with industry representatives. Each SIC code has four digits. Special surveys often add a fifth digit or letter to better identify specific groups of products. A new six-digit North American Industry Classification System (NAICS) is being introduced beginning in 1998. Its primary goal is improved relevance to current economic activity, in response to the rapid creation of new industries and the need to better characterize existing industries. Additional goals are to include Canada and Mexico in a unified system and to facilitate comparison of economic activity in North America with activity in the rest of the world. Classification under NAICS is based primarily on production methods; that is, products that use identical or similar production processes will be grouped together (Manufacturing News, 1997). A critical issue for all these data sources is that the standard system of industrial classification (see Box 6.3) does not align well with the various segments of the optics industry. This makes it extremely difficult to compare data from different sources, draw inferences about trends, or assess the effects of policies. Even though it produces billions of dollars in revenues each year, there is no statistically well-defined "optics industry." Many important and growing components of the optics industry described in this report cannot be found in the Standard Industrial Classification (SIC) database. For example, the category laser systems and equipment, except communications (SIC code 36992) is included only as part of electrical machinery, equipment, and supplies not elsewhere classified (SIC code 3699), which also includes Christmas tree lights, electric insect lamps, automatic garage door openers, and outboard electric motors. There is no way, except through a special survey, to sort out the data for laser systems, and the special survey that identifies laser systems and equipment, except communications is performed only every 5 years. Despite recent efforts by the Department of Commerce to provide some much-needed increased visibility to the rapidly growing high-technology segments of the U.S. industrial base, much of the optics industry appears to have been overlooked. For example, the new North American Industry Classification System (NAICS) appears to have dropped the category laser systems altogether! The Bureau of the Census should involve representatives of the optics industry in the next revision of the NAICS codes.
OCR for page 272
Harnessing Light: Optical Science and Engineering for the 21st Century TABLE 6.4Some NAICS and SIC Codes for Optics-related Products NAICS Description SIC Description 333314 Optical instruments and lens manufacturing 3827 Optical instruments and lenses 333315 Photography and photocopy equipment manufacturing 3861 Photography and photocopy equipment 334413 Semiconductor and related device manufacturing 3674 Semiconductor and related devices 334613 Magnetic and optical recording media manufacturing 3695 Magnetic and optical recording media 33511 Electric lamp bulb and part manufacturing 3641 Electric lamp bulbs and parts 335921 Fiber-optic cable manufacturing 3357 Nonferrous wire drawing and insulated wire and cable 336321 Vehicular lighting equipment manufacturing 3647 Vehicular lighting equipment 339112 Surgical and medical instrument manufacturing 3841 Surgical and medical instruments 339117 Eyeglasses and contact lens manufacturing 5995 Eyeglasses and contact lenses Some optics-related NAICS and SIC codes can be found in Table 6.4. In 1997 the major optics-related professional and trade organizations have come together to form the Coalition for Photonics and Optics (CPO). The goal of the CPO is to better represent the character of the optical industry as a unified whole. As the CPO develops, it is to be hoped that more (and more consistent) details about the character of the industry will become available. Summary and Recommendations Small companies, generally quite specialized, make up much of the U.S. optics industry. Thus, although the industry makes a significant contribution to the economy, this contribution comes in so many small pieces that it is hard to fully recognize and understand. Annual revenues are in the tens of billions of dollars, but precise characterization is difficult because the optics industry does not align well with the government's standard statistical categories. Furthermore, no single professional or trade organization represents the entire industry, although in recent months several major organizations have come together to form the Coalition for Photonics and Optics. New technologies are presenting new manufacturing challenges. Aspheric and diffractive elements offer new options for design and packaging, but design capabilities in this area currently far exceed capabilities for economical manufacturing. Driven by the growth of broadband telecommunications and other applications, fiber devices
OCR for page 273
Harnessing Light: Optical Science and Engineering for the 21st Century are developing rapidly and demanding lower-cost manufacturing techniques. New opportunities are emerging for semiconductor devices as their cost per watt falls. The photonic materials InP, GaAs, and LiNbO3 are maturing, but for future devices, CdTe, SiC, GaN, and AlN are still in need of improvements in crystal growth and material quality. Optical design is a key strength of the U.S. optics industry. Design and analysis capabilities have become dramatically more powerful as a result of advances in computing power. For photonic components, however, CAD tools like those used in the silicon industry are not available. Such tools are becoming essential for shortening design cycles and increasing functionality. The Department of Defense has historically been a key customer of the U.S. optics industry, and its special needs have strongly influenced the industry's development. Despite the trend toward increased reliance on commercial, off-the-shelf technology, continued DOD support of optics manufacturing technology is vital. There are many specialized optical technologies with important military applications whose commercial markets are insufficient to motivate development by the private sector alone. Photolithography equipment is a key comercial application of high-value specialty optics. Achieving the ever-smaller feature sizes demanded by the semiconductor industry will require industry support for improved fabrication of deep-UV aspheric elements. Metrology may be the single most difficult challenge in manufacturing the next generation of photolithography equipment. Photonics applications in information technology are a major driver of the optics industry's low-cost mass markets. Today, most photonic devices consist of discrete components. Advances will depend on lower cost and increased functionality, and these requirements will drive integration. Also required will be high-volume automated manufacturing lines; improvements in assembly and process control; and well-defined standards for form, fit, and function. Packaging, testing, and fiber connection account for the majority of the cost of photonic components; skills in these areas are underemphasized in the U.S. optics community. The shrinking pool of U.S. talent in bulk crystal growth is also an area of concern. Successful competition in international markets depends on establishing workable international standards. Virtually all new optics standards have hitherto been developed overseas, with no support from U.S. industry or the U.S. government. As a result, the U.S. optics industry has been a follower, not a leader, in adapting to new international opportunities. Active government and industrial participation in setting strong international standards for optical components is especially important because of the diversity of the U.S. optics industry.
OCR for page 274
Harnessing Light: Optical Science and Engineering for the 21st Century As discussed earlier in this chapter, the committee recommends the following actions: Government agencies and the optics community should recognize the importance of optics standards, especially their significance in international trade. The U.S. government should participate actively in the setting of such standards. NIST should be given the funding necessary to take the lead in this area. Government agencies should continue to support the activities necessary to introduce cost-effective precision aspheric components into both military and commercial products. DOD should continue to maintain technology assets and critical skills in optics manufacturing in order to meet future needs. The Bureau of the Census should involve representatives of the optics industry in the next revision of the NAICS codes. Collaborative programs in optics manufacturing should include universities so that students are trained in the latest technical solutions to production problems. DOD, NIST, and the DOE national laboratories should establish together a cooperative program that provides incentives and opportunities to develop new ideas into functioning methods for optics fabrication. References Belforte, D. 1997. Annual market review. Industrial Laser Review (January):17-21. Business Communications Company. 1996. The Future for Optical Coatings, RGB-187. Norwalk, Conn.: Business Communications. Council on Competitiveness. 1996. Endless Frontier, Limited Resources. Washington, D.C. Joint Precision Optics Technical Group. 1987. Precision Optics Study. Washington, D.C.: Pentagon Joint Group on the Industrial Base. Manufacturing News. 1997. Goodbye SIC, Hello NAICS. January, p. 1. The Photonics Directory. 1997. Pittsfield, Mass.: Laurin Publishing Company. U.S. Bureau of the Census. 1997. 1995 Annual Survey of Manufactures, M95(AS)-2. Washington, D.C.: U.S. Government Printing Office. U.S. Bureau of Export Administration (BEA). 1994. Critical Technology Assessment of the U.S. Optoelectronics Industry, NTIS PB93-192425. Washington, D.C.
Representative terms from entire chapter: