Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 Lighting Susan Walsh Sanderson Rensselaer Polytechnic Institute Kenneth L. Simons Rensselaer Polytechnic Institute Judith L. Walls University of Michigan Yin-Yi Lai Rensselaer Polytechnic Institute INTRODUCTION Once a symbol of Edisonâs creative genius and the prowess of American innovation, the incandescent light bulb represents a mature technology, now mas- tered by new competitors and imported at pennies apiece from China. Lamp (the industry name for a light bulb) manufacturing was dominated for decades by a few firms, notably Philips, OSRAM, and General Electric (GE). Related industry segments have typically been more fragmented, with thousands of firms produc- ing fixtures ranging from simple sconces to elaborate chandeliers. Increasingly both lamp and fixture manufacturing have been shifting to offshore locations, primarily in Asia. Not only are North American and European lamp and fixture companies under the threat from low-cost imports, but solid-state lighting, a semiconduc- tor- instead of bulb-based technology with greater potential energy efficiency and new capabilities, is poised to revolutionize the industry and change how we understand and use lightingâa change that will affect both traditional lamp and fixture producers. Solid-state lighting is challenging incumbents and throwing leadership in the future industry up for grabs. As innovative products composed of light emitting diodes (LEDs) are developed, new features like colors that change on command are expanding architectural possibilities. Other opportunities come from the convergence of lighting, information, and display technologies. In fiber optics light is data, and ordinary flat panel indoor lighting can serve as data transfer hubs, sending information to computers and appliances. Edisonâs lamp, and its successors, may soon be replaced with glowing ceiling panels or even lighting-enhanced wallpaper that changes patterns on command. 163
164 INNOVATION IN GLOBAL INDUSTRIES Which firms will successfully ride this new wave of innovation and what impact these changes will have on incumbents are not yet determined. Although the first wave of lighting innovation in the early 20th century spawned the devel- opment of global companies like GE, OSRAM, and Philips, these 21st-century innovations will create challenges for incumbents. New firms are emerging at all levels of the value chain to address the opportunity presented by solid-state lighting technologies. In this chapter we contrast traditional lighting technologies with LED technologies. Traditional lighting technologies we define as incandescent, gas- discharge, and electric arc lighting (which includes fluorescent, high-intensity discharge, mercury and sodium vapor, metal halide, and neon lamps). We exclude lighting technologies such as chemiluminescence that yield insufficient light for illumination (such lights can be seen but not seen by). LED technologies (includ- ing organic and polymer LEDs) are the only nontraditional technology considered because LEDs are the only alternative lighting approach that has reached suf- ficient maturity to be considered commercially viable in the trade, technology, and technical literatures. This chapter analyzes changes in lighting technology over the past two decades and its implications for U.S. industry competitiveness. We explore whether the rise of global competition is limited to low-cost manufacturing or whether strategic centers of decision making and research are moving away from the regions and firms that once dominated the industry. We examine the causes of these changes and what aspects of innovation in lighting, particularly in the arena of research and development (R&D), have changed since the 1990s. We speculate about the implications of these changes for firm strategy in the new era of intense global competition, we analyze how national policies have affected the development and diffusion of traditional and new lighting technologies, and we explore how public policy can best address the challenges and opportunities of- fered by solid-state lighting to aid countries in their struggles to conserve energy and reduce global warming. We are entering an era of faster-paced competition as the lighting industry, which has been dominated by a few firms (at least in the lamp sector), faces com- petition from new technologies, firms, and regions. Asian firms, as well as firms headquartered in the United States and Europe, have performed strongly in patent invention for solid-state lighting and are making key contributions to these new technologies. Both new firms and incumbents are investing heavily in solid-state lighting technologies, and it remains to be seen which firms will predominate. Public policy will likely play an important role in future developments by stimulating demand for energy-saving lighting, providing funding for R&D, and incubating startup companies as they seek to commercialize these new technolo- gies. But retail firms like Wal-Mart are increasingly playing a role in the diffusion of energy-saving lighting technologies. We compare the policies of countries
LIGHTING 165 supporting development and diffusion of new lighting technologies and specu- late about how these efforts may affect the location of R&D, manufacturing, and headquarters of surviving lighting producers. EVOLUTION OF THE LIGHTING INDUSTRY Globalization of Lighting Production The global lighting market in 2004 was worth some $40 billion to $100 bil- lion, about one-third of which represented lamps. U.S. apparent consumption of lamps, fixtures, and other equipment totaled about $14.8 billion in 2004. U.S. production of lamps grew steadily until the early 1970s, then fluctuated over the next 20 years, stabilized during the 1990s at about 1970 levels, and finally fell somewhat at the start of the 21st century, as shown in Figure 1. The eventual leveling off and downturn in U.S. lamp production in the 1990s can be explained, in part, by a steady increase in imports over the past two de- cades. Total imports, as a percentage of U.S. apparent consumption, increased from less than 20 percent in 1989 to around 50 percent in 2004, as shown in Figure 2. About half of the imports come from China, Mexico, and Japan, with China representing the largest share as of 2004. In 1989, less than 3 percent of lamps were imported from China. By 2004, Chinese lamp imports represented 26 percent of all lamp imports, having grown more rapidly than imports from any other supplier nation, and 10 percent of apparent lamp consumption in the United States. Once concentrated in the hands of three large manufacturers, the incan- descent bulb industry has new competitors, primarily low-cost manufacturers in Asia. In the fixtures industry, broken down in Figure 2, these trends are more intense, with 86 percent of all fixtures imports in the United States arriving from China by 2004. Increased fixture imports are the result of both incursion of lower-cost Chinese manufacturers and shifting production abroad by U.S. firms that seek lower-cost manufacturing sites. An exception to this trend is Genlyte â Hadley et al. (2004) cite the figure of $40 billion, one-third of which represents lamps, but pub- licly available estimates of the size of the global lighting industry vary greatly, and the firm Color Kinetics in a communication with us cites the figure of $100 billion based on data from Fredonia Marketing Research. â Apparent consumption equals U.S. production plus imports less exports, where U.S. production is measured as value of shipments from Bureau of Economic Analysis data, and import and export data are from the U.S. International Trade Commission. â Not shown in Figure 2 is an additional trade category âOther Lighting Equipment,â for which im- ports increased from 38 percent in 1989 to 57 percent in 2004, with Chinaâs share of imports growing from 24 percent in 1996 to 32 percent in 2004.
166 INNOVATION IN GLOBAL INDUSTRIES $mil $4,500 $4,000 $3,500 $3,000 $2,500 $2,000 $1,500 $1,000 Lamps Residential Fixtures $500 Other Lighting Equipment $0 2000 2004 2002 1960 1964 1966 1968 1990 1996 1980 1984 1986 1988 1998 1962 1994 1958 1982 1992 1972 1978 1970 1976 1974 FIGURE 1â U.S. shipments of lighting products (real 2004 values). SOURCES: Ship- lighting-1.eps ment values: National Bureau of Economic Research (1958-1996), U.S. Department of Commerce, U.S. Census Bureau (1997-2001), Bureau of Economic Analysis (2002-2004). Producer Price Index from the Bureau of Labor Statistics. Thomas, the largest lighting fixture and control company in North America, which manufactures 70 percent of its products in the North American region in order to keep close to its design centers and customers (Genlyte Thomas, 2005). Genlyte Thomas is introducing new energy-efficient light fixtures using compact fluorescent (CFL), high-intensity discharge (HID), and LED lamps and is con- ducting research on solid-state lighting to remain the premier fixtures company while the industry transitions to new lighting technologies. The remaining area of growth for U.S. lighting production in the 1990s was in specialty lighting applications, such as Christmas decorations, underwater lighting, and infrared and ultraviolet (UV) lamps. This sector grew steadily throughout the second half of the 20th century and, as Figure 1 reveals, has sur- passed the production value of lamps and of residential fixtures. â Figures 1 and 2 use the definition of traditional lighting, as defined in the beginning of this article, for âlampsâ based on SIC (3641, 3648) and NAICS (33511, 335129), which includes all traditional lamp types including (regular and compact) fluorescent and HID lamps, but excludes LED lamps.
LIGHTING 167 60% 50% Fixtures Imports 40% Fixtures Imports from China 30% Lamp Imports 20% Lamp Exports 10% Lamp Imports from China Fixtures Exports 0% 2000 2004 2002 2003 2001 1990 1996 1998 1999 1989 1994 1995 1992 1993 1997 1991 FIGURE 2â U.S. imports and exports of lamps and fixtures, total and imports from China, as percentages of U.S. consumption. SOURCE: U.S. International Trade Commission. lighting-2.eps Big Three Lamp Producers While there are hundreds of small lamp producers, which usually specialize in one type of lamp, the global lamp market is dominated by three big players: Koninklijke Philips Electronics (Philips), OSRAM-Sylvania (OSRAM), and GE. All three firms produce a wide spectrum of lamps based on distinct technolo- gies for most major commercial and residential markets. Philips has the largest global market share in lamps, and GE has the largest U.S. market share (Mintel, 2003). In the United States, GE has been a dominant player in lighting since the industryâs inception (Leonard, 1992). As early as the mid-1890s, GE and West- inghouse controlled a 75 percent market share. GE eventually gained even greater market dominance, so that by 1927 GE and its licensees held 97 percent of the U.S. lamp market. Hygrade-Sylvania, whose lighting operations would much â That most lamp producers specialize on a single type of lamp is apparent from industry directories such as www.lightsearch.com. â Philips Lighting employs about 45,500 people and has 70 manufacturing facilities worldwide (Philips, 2006, p. 38).
168 INNOVATION IN GLOBAL INDUSTRIES later be bought by a German producer to form OSRAM-Sylvania, was GEâs larg- est lamp licensee. Although GEâs market dominance fell in the latter half of the 20th century, it remained the largest U.S. lamp producer. In Europe, the lamp market also became concentrated early (Leonard, 1992). The leading firm was OSRAM, formed in a 1919 merger of the three leading German lamp producers, and now wholly owned by Siemens. Second in the Eu- ropean market was the Dutch company Philips. In part through cooperation with a European cartel set up in the 1930s under Swiss corporation Phoebus S.A., GE made substantial inroads in Europe and became the dominant worldwide pro- ducer. The big three lighting firms all maintained leading positions in traditional lighting technology. Traditional electric lighting patent applications during the period 1990-1993 were identified using data for the United States and Western Europe. As noted earlier, we define traditional lighting to include incandescent, gas-discharge, and electric arc lighting (which includes fluorescent, HID, mercury and sodium vapor, metal halide, and neon lamps). All of the big three were leaders in these traditional electric lighting technologies, with 257.8 patent applications by Phil- ips (credit is split equally in the case of multiple assignees); 232.1 applications by GE and by Thorn, whose lighting business GE acquired in 1991; and 219.4 applications by OSRAM, Sylvania, and OSRAMâs parent firm Siemens. The big three each had more patent applications than any other firm. â GEâs dominance varied substantially across nations. For example, in the United Kingdom in 1965-1967, the leading producer was British Lighting Industries, followed by Philips and OSRAM (Monopolies Commission, 1968, p. 8). â Patents are included for international patent classifications H01J61-65, âDischarge lampsâ; H01K, âElectric incandescent lampsâ; and H5B31 and H5B35-43, which cover âElectric lighting . . . not otherwise provided forâ excluding electroluminescent light sources (which provide sufficient light to see an object but not to see by). Patents are included for applications at patent offices of the United States (1,589 applications), Europe (976), Austria (190), Belgium (20), Denmark (49), Spain (976), Finland (79), France (121), Germany (1,798), Ireland (3), Italy (31), Netherlands (51), Norway (22), Portugal (5), Spain (194), Sweden (20), Switzerland (22), and the United Kingdom (218). Data are drawn from the European Patent Officeâs Worldwide Patent Statistics Database, version April 2006 (with the coverage of the Espacenet online database). Equivalent applications in multiple nations, detected by the fact that they share an identical set of priorities (as in Espacenet), were treated as a single application by weighting each application in inverse proportion to its number of equivalents (including itself). â German patent trust, Patra Patent Treuhand (possibly associated with OSRAM or Siemens), A had 185.7 applications. The next three firms in number of patent applications were Toshiba with 70.2 applications, Motorola with 36 applications, and Matsushita with 33 applications. As in most areas of patenting, there were patent applications by many other individuals and companies (the total number of relevant patent applications was 3,236 during the period 1990-1993), and meaningful analyses are based on relative numbers, not on percentages of total applications. When figures are measured in terms of the number of patents actually granted from these applications (by the time of data collection), the conclusions are similar: Philips received 213.6 patents; OSRAM, Sylvania, and Siemens, 205; GE including Thorn, 181.5; Patra Patent Treuhand, 76.9; Toshiba, 52.6; Motorola, 35; and Matsushita, 31.
LIGHTING 169 Evolving Technology The lighting industry has developed several types of lamps. The incandes- cent lamp, little changed in form since the Edison era, is an evacuated glass tube (usually refilled with a gas) in which an electric current passes through a thin filament, heating it and causing it to emit light. Mercury vapor lamps, first patented in 1901 by Peter Cooper Hewitt, are high-pressure gas arc lamps and a forerunner to fluorescent lamps. Neon lamps were invented by Georges Claude 10 years later. Fluorescent lamps, first patented by Meyer, Spanner, and Germer in 1927, use a glowing phosphor coating instead of glowing wires to increase ef- ficiency. Special types of incandescent lamps, such as bulbs filled with halogen gas to increase lifetime and efficiency, have also been developed. Incandescent lamps account for a majority of household sales in the United States, but a smaller portion of total sales. In households, incandescent lamps represent 66.5 percent of sales revenues, whereas fluorescent and other lamps re- main uncommon (Mintel, 2003). Residential sales, however, make up less than 10 percent of lighting demand measured in lumen-hours. Combining all economic sectors (residential, commercial, industrial, and outdoor), incandescent lamps represent 11.0 percent of lumen-hours of light output, as compared to about 57.5 percent for fluorescent, 31.0 percent for HID, and 0.01 percent for solid-state lighting (Navigant Consulting, 2003a, p. 7). Each of these lamp types has experienced a steady march of small improve- ments in materials, design, light quality, energy efficiency, and manufacturing efficiency throughout the past century. While early improvements were made by independent inventors in the United Kingdom, more than three-quarters of these improvements originated in countries where the big three were headquar- teredâthe United States, the Netherlands, and Germany.10 In materials, for example, thorium oxide added to wires increased shock resistance, nonsag wire formulations made possible new configurations for brighter and more easily mounted incandescent filaments, and safer phosphors replaced the highly toxic beryllium coating in fluorescent lights. Examples of design changes include fill- ing incandescent lamps with large-molecule gases to prolong filament lifetimes, new layouts of filament mounts to facilitate assembly and automated manu- facture, and a proliferation of lamp varieties, shapes, and sizes. Light-quality changes were achieved by choosing appropriate filament and phosphor materials and sometimes by blocking part of the emitted light to attain, for instance, a look similar to sunlight. Energy-saving lamps also progressed steadily but slowly. GE, for example, commercially introduced its first energy-saving incandescent lamp in 1913, but 10â We catalogued 134 improvements in lamp technologies between 1905 and 2005. Sources: com- pany websites of General Electric (2006), OSRAM (2006), Siemens (2006), Philips (2006), and Toshiba (2006); websites of Bellis (2006a,b,c), Williams (2005), and Arthur (2006), and Bowers (1982).
170 INNOVATION IN GLOBAL INDUSTRIES not until 1974 was the first energy-saving fluorescent lamp introduced. Manu- facturing became increasingly efficient with machines and methods that allowed faster, higher-quality production with less manual labor. Automatic insertion and mounting of components, sealing, exhausting, basing, and flashing were key process technologies. Many of these and other improvements took place during the first half of the last century and are documented by Bright (1958, pp. 22-30). In the latter half of the century, improvements focused largely on improved ef- ficiency and longer lamp lives. The discovery of substances such as narrowband phosphors led to the development of CFLs, gases such as xenon yielded brighter lamps such as those used in automobiles, and similar improvements had medical uses including UV lamps. Whereas lowering manufacturing costs and streamlining production were the key lighting challenges of the late 20th century, saving energy is the new driv- ing force for 21st-century development. Lighting accounted for about 22 percent of total energy used in residential and commercial sectors in the mid-1990s, as shown in Figures 3 and 4 (DOE, 1995, 1997). In 2001, 51 percent of the national energy consumption for lighting occurred in the commercial sector, 27 percent in residences, and 14 percent in industry; the remaining 8 percent was used in outdoor stationary lighting (Hong et al., 2005, p. 2). Almost half of the electricity used in commercial buildings is used in lighting, as Figure 5 indicates. In the United States, residential homes largely use incandescent lamps (90 percent), whereas commercial and industrial sectors use mostly fluorescents (Hong et al., 2005). If residential homes in the United States replaced all incan- Refrigera- tion Ventilation Cooking 3% 3% 4% Office Lighting Equipment 23% 6% Cooling 7% Other 7% Water Space heating heating 15% 32% FIGURE 3â Energy consumption in U.S. commercial sector, 1995. SOURCE: DOE (1995). Lighting fig 3 NEW
LIGHTING 171 Aircon Refrigera- Lighting & 4% tion other 4% appliances 22% Water heating 19% Space heating 51% FIGURE 4â Energy consumption in U.S. residential sector, 1997. SOURCE: DOE (1997). Lighting 4 new descent lamps with CFLs, they would save an estimated 35 percent of electricity used for all lighting applications (DOE, 1993). Although advances in energy-saving lighting technologies such as CFL have been an important part of the strategies of the big three lamp producers, the big three have had some difficulty getting residential customers to give up incan- descent bulbs and replace them with the more energy-efficient but initially more expensive bulbs. The rate of adoption of CFLs in U.S. residential households has been low, particularly compared to that of Europe and Asia. Researchers at- tribute those differences to a variety of factors, including national coordination of promotional efforts, different cultural attitudes about resource consumption, and higher electricity prices (Calwell et al., 1999). U.S. residential consumers lack awareness of and knowledge about CFLs. Consumer buying habits, negative perceptions, and skepticism about fluorescent lighting and relatively low electric- ity prices have meant that the United States is behind the rest of the world in adoption of energy-saving lighting technologies (Sandahl et al., 2006). This may soon change; for example, Wal-Mart CEO H. Lee Scott, Jr., is committed to sell 100 million CFLs a year by 2008 and the firm is making a concerted effort to change consumer behavior (Barbaro, 2007).11 11â Wal-Mart sold about 40 million CFLs compared to 350 million incandescent light bulbs in 2005.
172 INNOVATION IN GLOBAL INDUSTRIES Water Space heating heating 2% Cooking 4% Ventilation 1% 6% Refrigera- tion 7% Lighting Other 46% 8% Office Equipment 13% Cooling 13% FIGURE 5â Electricity consumption in U.S. commercial sector, 1995. SOURCE: DOE (1995). Lighting 5--NEW Since lamp efficacy is central to which lamp types dominate the market, it is important to understand efficacy and its role in purchasing decisions. Efficacy in lighting can be measured in terms of lumens produced per watt of electricity (lm/W). A standard 100-watt incandescent lamp, for example, lasts about 1,000 hours and produces 15 lm/W. By comparison, a standard 30-watt fluorescent lamp lasts 20,000 hours and produces 80 lm/W. A longer-lasting and more energy-ef- ficient bulb is less costly over the long term but higher initial upfront costs and misconceptions about the efficacy of fluorescent lights (early fluorescents had poor color rendering and were noisy) have led to low adoption in residences. Op- timal lamp choice involves not only energy efficiency but also replacement costs for burned-out lamps and labor costs to install lighting systems. In commercial and industrial settings, where life-cycle costs are important and companies can make upfront investments, fluorescents are usually chosen. RADICAL INNOVATION IN LIGHTING: LEDS Nature and Advantages of LEDs An LED is a semiconductor diode. It is electroluminescent, emitting color that depends on the chemical composition of the semiconductor material or com- pound used and ranges along the spectrum from UV to infrared, as documented in Table 1.
LIGHTING 173 TABLE 1â LED Color Spectrum Available from Alternative Materials Semiconductor Material Color AlGaAs (aluminum gallium arsenide) Red Infrared AlGaP (aluminum gallium phosphide) Green AlGaInP (aluminum gallium indium phosphide) Orange-red (bright) Orange Yellow Green GaAsP (gallium arsenide phosphide) Red Orange-red Orange Yellow GaP (gallium phosphide) Red Yellow Green GaN (gallium nitride) Green Pure green (emerald) Blue InGaN (indium gallium nitride) Bluish green Blue Near UV SiC (silicon carbide) as substrate Blue Si (silicon) as substrate, under development Al2O3 (sapphire) as substrate ZnSe (zinc selenide) C (diamond) UV AlN (aluminum nitride) Far UV AlGaN (aluminum gallium nitride) SOURCE: Wikipedia (2006). The first practical visible-spectrum LED was developed in 1962 by Nick Holonyak (Inquirer, 2004), and a variety of single-color LEDs followed. White LEDs have been a long-standing goal for researchers since they are most likely to replace traditional bulbs. White LEDs have been created by coating blue LEDs with a yellow phosphor, yielding a blue and yellow glow that appears white to the human eye. Another approach, taken by GE, uses UV LEDs driving phosphors, and a third approach is to use multiple colors of LEDs and combine them to create white light. Current white LEDs are cost-effective only for certain applications, such as backlighting and flashlights, and color LEDs remain more widely used. Although incandescent and fluorescent lamps remain the predominant light
174 INNOVATION IN GLOBAL INDUSTRIES sources, LEDs have several potential advantages. First, they use less energy: LEDs are three- to fourfold more efficient than incandescent and halogen sources. However, with the exception of laboratory devices, LEDs still fall short of fluorescent sources for many white light applications. Nevertheless, they are semiconductor devices and LED lighting is thought to follow an equivalent of Mooreâs law in computing, advancing rapidly and continually because of the pace of electronic circuit improvements. Second, in contrast to incandescent lamps, LEDs use most of their energy in lighting (Herkelrath et al., 2005). LEDs also have a long life span, typically about 10 years of on-timeâtwice that of fluorescent lamps and twenty times that of incandescent lamps. In terms of luminous efficacy (lm/W), LEDs by 2004 were already about four times as efficient as incandescent lamps, and by 2020 they are targeted to be about 12 times as efficient as current incandescent lamps and more than twice as efficient as current fluorescent lamps (Hadley et al., 2004, p. 5; Tsao, 2002, p. 4). In addition, LEDs light up many orders of magnitude faster than incandescent lamps, and, rather than burning out abruptly, they do so slowly. LEDs require little maintenance and are cool to the touch, durable, and flexible. Furthermore, the technology is digitally compatible and, hence, can be integrated into digital networks, facilitating customizable electronic control. LEDs come in many shapes and sizes and have multiple uses. Backlighting is one use, for cell phones, cars and other electronics, liquid crystal displays (LCDs), and specialized lighting applications. Specialty uses are possible since LEDs can be waterproofed, bent, shaped, multicolored, and dimmed.12 LED applications are common in the entertainment industry, hotels, road signs, exit signs, pools, landscaping, and darkrooms. The main drawback of LEDs is that they have not yet achieved the efficacy necessary for many white light applications. They are also still costly because they are expensive to produce. But production costs are expected to decline as volumes rise and the technology advances. For example, in 2002, the total cost of LED lamps (capital cost plus operating costs) was estimated at $16.00 per million lumen hours, compared to $7.50 for incandescent bulbs and $1.35 for fluorescent bulbs (Hadley et al., 2004, p. 8; Tsao, 2002, p. 8). However, by 2020, the total (capital plus operating) cost of LED lamps is targeted to be reduced to $0.63 per million lumen hours (Hadley et al., 2004, p. 8; Tsao, 2002, p. 8). An additional limitation of LEDs, relative to incandescent lamps, is their im- perfect color rendering, given the spectrum of light emitted. White light created by multiple color LEDs or by phosphors driven by LEDs involves a combina- tion of wavelengths of light that differs from the color spectrum of traditional lamps and of sunshine, making objects with certain colors appear relatively dark. 12â LEDs can also be designed to trap insects (through the use of insect-attracting colors) or to avoid attracting insects (since they do not generate UV light) (Bishop et al., 2004).
LIGHTING 175 However, Ashdown et al. (2004, p. 8) indicate that color spectrum limitations are likely to be remedied as the technology progresses.13 Until recently, LEDs have been the only viable technology competing with the various types of traditional lamps as electricity-driven sources of illumination (Hong et al., 2005). A recently emerged technology that may become competitive with LEDs is microwave-driven light bulbs, for which claims of efficiency and low cost remain to be assessed; this technology emerged during the final stages of our work and is not assessed here (Economist, 2007a,b). Our analysis (including patent data presented later) includes two newer types of LEDs: organic light- emitting diodes (OLEDs) and polymer LEDs. OLEDs, which are LEDs involving organic (carbon-based) chemicals, are promising but still in an early develop- ment stage. Ching Tang and Steven Van Slyke of Eastman Kodak invented the first OLED in 1987 (Howard, 2004). The materials in OLED devices have broad emission spectra that provide an advantage over inorganic LEDs (minor changes in the chemical composition of the emissive structure can tune the emission peak of the device). It is believed that good-quality white light is achievable from OLEDs (OLLA, 2006a, 2006b).14 An important focus of current OLED research is on improving operational life. In particular, OLEDs are of interest to display firms since they are capable of producing true black colors, something LCDs cannot achieve since they require a backlight to function and are never truly âoff.â OLEDs can produce a greater range of colors, brightness, and viewing angles than LCDs because OLED pixels emit light directly. The display industry, with more than 70 companies includ- ing OLED pioneer Eastman Kodak, is set to commercialize OLED technology including OLED displays (Hong et al., 2005).15 Kodak launched the first digital camera to use a full-color OLED display in 2003. The big three traditional light- ing companies have all set up joint ventures to profit from OLED technology for the display market. 13â Other limitations of LEDs are areas of active work. For example, LEDs driven with sufficient power for automotive headlights and taillights require heat sinks because heat degrades LEDs; rel- evant heat sinks are improving. 14â White OLEDs by 2006 had achieved claimed efficacy of 25-64 lm/W (Burgess, 2006; Physorg. com, 2005; Ledsmagazine.com, 2006). 15â December 2007, Sony released the first commercially available OLED televisions, although In they remain far from price-competitive with liquid crystal display (LCD) televisions (Eisenberg, 2007). Samsung has also been demonstrating prototype OLED displays (Gizmodo, 2008). LCDs, which are now dominant in televisions and computer monitors, are akin to (dynamically changing) stained glass windows behind which white light is projected, thus allowing some colors of light through the liquid crystals while blocking other colors. This means that light not allowed through the liquid crystals is wasted, turning electricity into heat, and that the liquid crystals require behind them a light source that adds to their thickness, weight, and cost. If OLED displays and their production methods can be improved sufficiently, therefore, they have potential advantages over LED displays: the desired color can be produced at each location on an OLED display with no wasted light and no liquid crystal layer.
176 INNOVATION IN GLOBAL INDUSTRIES LEDs as a Disruptive Technology: Diffusion Among Applications Disruptive technology has been defined in several ways, and LEDs match at least two of the definitions. First, disruptive technology may be defined as a new technology that fills a long-standing need and for which the expertise and resources of incumbent firms do little to help them with this new approach; in this case solid-state lighting fills the need for illumination using a technology that dif- fers totally from traditional lighting technologies. Second, disruptive technology may be defined as a technology in which new firms enter a market and threaten the market dominance of incumbent firms; in this case new firms have been en- tering the lighting market by creating products based on LEDs, and it has been unclear whether the leading existing lighting manufacturers can maintain strong market positions if purchases shift substantially to LEDs. LEDs are a novel technology in lighting. LEDs are semiconductors and so manufacture of LEDs has little in common with traditional lamp production. The supply chain to produce LEDs as indicated in our discussions with industry ex- perts is quite disaggregated, as is the case for other semiconductor technologies. Moreover, Kevin Dowling, Chief Technology Officer at Color Kinetics, stated to us that âthe vertically integrated giants of the semiconductor world such as Intel and Applied Materials are becoming less numerous and rapidly becoming more the exception rather than the rule.â Data on the participation of firms in each stage of the LED supply chain are available from solidstatelighting.net (2006). We catalogued the participation of each firm in each stage of the supply chain and found that most firms participate in only a single part of the supply chain, although a few large firms are involved in many parts of the supply chain. This supply chain is illustrated in Figure 6. Semiconductor firms often specialize in specific stages of the supply chain, such as R&D, epitaxy, manufacturing, packaging, testing, and back-end pro- cessing. Each stage requires unique skills and equipment and significant capital expenditure, which is one reason why firms tend to specialize rather than inte- grate along the supply chain. Specialization is thought to drive down prices and improve performance, and this trend is similar among LED manufacturers. The development of this new technology will likely create opportunities at all levels of the value chain.16 The LED market in general lighting is still small compared to traditional lamps. LED applications command a high price, but relatively few units are sold and all are for specialty purposes. Traditional lamps (incandescent, fluorescent, and HID) are estimated by Navigant Consulting (2003a, p. 7) to have used 41,051 trillion lumen hours of electricity in the United States in 2005 compared to only 16â As lighting moves to LEDs, traditional bulb manufacture may also be vertically disintegrating somewhat, judging from recent comments by Jim Campbell, President and CEO of GE Consumer & Industrial (General Electric, 2007).
LIGHTING 177 Integrated circuit deposition removal patterning modification (growing, (etching) (lithography) (doping) epitaxy R&D coating) interconnect test box, label, (levels) ship Raw Wafers Wafer Test Back-end Arrays/ materials Assembly fabrication processing Fixtures/ (compounds) Systems Manufacturing Integrators grow slice polish score test ingots wafer cut wafer attach to bonding encapsulate into dies lead frame Packaging FIGURE 6â LED semiconductor supply chain. lighting-6.eps 5 for LEDs. Nevertheless, the LED market grew by 50 percent between 1995 and 2000 and has been forecast at $4.7 billion by 2007 (Ashdown et al., 2004, p. 9). LED technology has some clear advantages over traditional lighting, which have allowed LED manufacturers to displace traditional lighting in niche markets (Griffiths, 2006). Indicator lights were one of the earliest uses, with color LED indicators predating the 1990s and white LED indicators used from about 2000 onward. In 2001, 30 percent of LED sales were for backlighting, 26 percent for automotive uses, 26 percent for signs and displays, 10 percent for electronic equipment, 4 percent for signals, and 4 percent for general illumination (Mac- cagno, 2002, cited by Ashdown et al., 2004, p. 9). By 2002, U.S. market penetra- tion of LEDs was particularly high in exit signs (80 percent), truck and bus lights (41 percent), and traffic signals (30 percent) (Navigant Consulting, 2003b, p. xii; Hadley et al., 2004, p. 9). Other niches that LEDs have entered include video screen backlights in the mid-1990s, decorative lighting in the late 1990s, and automobile lights includ- ing dashboard, interior, brake, and tail lights since about 2000. Other recent uses include architectural lighting, outdoor advertising, and long-lasting white-light flashlights. One example of the advantage of LEDs is in brake lights, where LEDs provide an extra 0.2-second response time and therefore help to prevent acci- dents. Although first introduced in luxury cars, LED brake lights are beginning to penetrate the more cost-conscious end of the market. Another example is Wal-
178 INNOVATION IN GLOBAL INDUSTRIES Martâs adoption of LED lighting for refrigerated display cases, an application once dominated by fluorescents. As well as lowering operating costs, the LED lights are amenable to added motion sensors so that lights come on only when shoppers are nearby. As a result, Wal-Mart is investing $30 million and expects a 66 percent reduction in freezer lighting energy costs (LIGHTimes, 2006). 17 New LED Lighting Innovators Advances in solid-state lighting offer an opportunity and a challenge for incumbent and startup firms. Although the big three lamp manufacturers have been making substantial investments in solid-state lighting, pioneering inventors in LED lighting come from universities, research labs, and companies, and R&D plays a vital role in development of these technologies. Advances in red, yellow, and blue LEDs have been led by different research groups and companies. Several companies have âspecializationsâ in one industry sector, due to a combination of strategy and luck in pioneering key product or process innovations. Nichia Corp. in Japan, for example, was one of the first companies to develop blue LEDs, a key advance when only red, green, and yel- low were available. It also produced the first white LEDs in 1996 (Walker, 2004). The company is an illustration of how small firms have been able to penetrate the burgeoning industry. Nichiaâs key researcher, Shuji Nakamura, now a professor at the University of California, Santa Barbara, was largely responsible for the development of the blue LED. When Nakamura was hired in 1979, Nichia was a small firm in rural Japan with only 200 employees and Nakamura was assigned a project to synthesize a commercial-grade blue LED, needed to complete the color palette. At the time, large Japanese corporations were spending $85 million a year and Nichia had a very small research budget. Today Nichia controls 80 percent of the blue LED market with Cree and Toyoda Gosei (Cox, 2003). Nakamuraâs successful approach departed from the standard thinking in his field and in his company. He chose gallium nitride, a material most researchers thought would not yield significant results, as the basis for his research, and continued to work on blue LEDs for 10 years. Nichiaâs entry into the LED market was a lucky out- come of their hiring a particular employee and of that employeeâs actions. The role of individuals in innovative companies in pioneering new lighting technologies is typical of the early stages of a technology cycle in which R&D efforts are lengthy and costly. Government grants have been instrumental to sup- port startup companies and university research in the pursuit of emerging LED technologies. Government funding has filled key technology gaps, provided fund- 17â Wal-Mart installations developed in collaboration with GE and Philips represent âthe biggest investment to date in LED lighting for interior application [$30 million], and it is also the single largest installation of white LED lighting replacing fluorescent lighting in a display lighting applica- tionâ (Griffiths, 2006).
LIGHTING 179 ing to develop enabling knowledge and data, and advanced the solid-state lighting technology base. A team of researchers from Rensselaer Polytechnic Institute, for example, recently received $1.8 million in federal funding from the Depart- ment of Energy (DOE) to improve the energy efficiency of green LEDs, with a goal of doubling or tripling power output. The research was one of 16 projects selected for funding through the DOEâs Solid-State Lighting Core Technologies Funding Opportunity Announcement, which supports enabling and fundamental solid-state lighting technology for general illumination. âMaking lighting more efficient is one of the biggest challenges we face,â says Christian Wetzel, the Wellfleet Career Development Constellation Profes- sor, Future Chips, and associate professor of physics at Rensselaer (RPI, 2006). To meet aggressive DOE performance targets that call for more energy-efficient, longer-lasting, and cost-competitive solid-state lighting by 2025, the team has partnered with startups such as Kyma Technologies and Crystal IS. Kyma, a North Carolina State University spin-off, specializes in gallium nitride substrates, while Crystal IS, founded by two Rensselaer professors, specializes in blue and UV lasers based on single-crystal aluminum nitride substrates. The DOE grant has funded these startups and researchers. Government support has also been important for building demand and aid- ing firms to improve quality and reduce pricesâkeys to further diffusion. Such programs promote early diffusion of energy-saving technologies and are not unique to the United States. We will return to the role of national policies and government initiatives later in the chapter. CORPORATE STRATEGIES TOWARD INNOVATION The âBig Threeâ The big three traditional lighting manufacturers, Philips, OSRAM, and GE, have responded to the opportunities in LED lighting by creating joint ventures with semiconductor firms that had preexisting expertise in these technologies. They later acquired these joint ventures outright. Philips established a joint venture in optoelectronics with Hewlett-Packard (HP) in 1999 (when HP split in two in 1999, the optoelectronics group was assigned to a new firm, Agilent Technologies) and acquired the venture Lumileds in 2005 for $950 million. 18 OSRAM established a joint venture with Infineon Technologies AG (formerly Siemens Semiconductors) in 1999, and acquired the venture in 2001, naming it OSRAM Opto Semiconductors GmbH.19 GE established a joint venture, Gelcore, 18â Philipsâ 2005 Annual Report states that Lumileds is the worldâs leading manufacturer of high- power LEDs. 19â Siemens gradually spun off its semiconductor division as Infineon beginning in 1999, and sold its final 18.23 percent stake in the company in 2006.
180 INNOVATION IN GLOBAL INDUSTRIES with semiconductor maker Emcore in 1999, and acquired Gelcore in August 2006 for $100 million.20 Additionally, Philips and OSRAM announced in January 2007 a cross-licensing agreement covering patents on LEDs and OLEDs (LIGHTimes, 2007), and Philips acquired the LED lighting controls firm Color Kinetics in August 2007. Historically, the locus of innovation for traditional (i.e., non-LED) lamps originated in the primary R&D centers of the big three lighting firms in Germany, the Netherlands, and the United States. While these labs are still very important, in recognition of Asiaâs increasingly important role in the traditional lighting industry, the three firms have set up manufacturing, engineering, and R&D activ- ity in other parts of the world, principally in Japan and Taiwan. Efforts are also being made to penetrate the Chinese market. GE, for example, began investing in China through joint ventures. The company combined a finished product purchasing center and an R&D center to form the GE Asia Lighting Center in Shanghai. By 2002, GE had four major plants in Shanghai and one in Xiamen, and had invested over $100 million in China for lighting, according to Matthew Espe, former president and CEO of GE Lighting (Zou, 2002). Philips established an R&D campus with the Shanghai Science and Tech- nology Commission with annual expenditures of $50 million, the majority of which is in lighting. Between 1988 and 2005, Philips Lighting established nine solely owned and joint ventures and five R&D centers, of which one conducts global level research while the other four mainly focus on the Asia-Pacific region (Chinesewings, 2005). OSRAM China Lighting, Inc., 90 percent of which is owned by OSRAM, was formed in April 1995 with an investment of 49.7 million Euros. The company is located in Foshan, China, and has two factories in China, employing 6,000. In February 2006, OSRAM China Lighting announced it would acquire Foshan Electrical and Lighting Co. Ltd.21 Asian LED Producers Beyond their expanding importance in the traditional lighting industry, Asian firms also play a significant role in the global LED market. Japanâs LED industry leads with $918 million in sales, or a world market share of 47 percent, although a portion of these revenues are shared with some U.S. companies through joint ventures. Taiwanâs industry holds second place at $712 million, or a market share of 25 percent (Taiwan Economic News, 2004a,b). LEDs are the largest type of 20â Although Gelcore grew about 50 percent from 1995 through 2004, it nonetheless reported a net loss of $0.8 million in 2005. Thus, when Emcore sold its 49 percent stake in Gelcore, it traded pos- sible future value for immediate cash gains. 21â Source: âOSRAM China Lightingâs Announcement to Acquire Foshan Electrical and Lighting Co. Ltd.,â http://business.sohu.com/20060614/n243727208.shtml (in Chinese).
LIGHTING 181 compound semiconductor production in Taiwan (Liu, 2003). Another source es- timates the global LED market at $5.4 billion in 2004, with Japanâs share at 51.3 percent, Taiwanâs 22.7 percent, the United States 12 percent, and Europe around 9 percent (Ledsmagazine.com, 2005b). Data from www.solidstatelighting.net (2006) suggest that most LED R&D is conducted in the United States whereas Asia dominates manufacturing and packaging. For example, Taiwan, China, and Korea produced the majority of blue LEDs in the world, and more than 80 percent of the production of InGaAlP high-brightness (HB) LEDs in 2004. China boasts about 600 enterprises directly related to the LED industry in China, employing about 40,000 people (Ledsmagazine.com, 2005a). Because of the youth of the industry, quality remains variable across firms and is sometimes difficult to assess. Quality for an LED includes a long operat- ing lifetime before substantial fading occurs, controls to ensure against defective chips, and consistency of operating characteristics including a claimed color emission spectrum, light intensity, voltage drop, and viewing angle (Toniolo, 2006b). The LED industry has seen a surge of new players, which has flooded the market with low-quality LEDs (Toniolo, 2006a). The result of such com- modity production is intense price competition and a âhuge overcapacityâ for lower-performance LEDs (Arensman, 2005). For high-performance LEDs, com- petition remains less intense. Among consumers, low-quality LEDs may give a bad reputation to all LEDs. New Ventures The global semiconductor market was worth $262.7 billion in 2006 (Gartner, 2007), considerably outstripping LED industry revenues of $3.7 billion in 2004 (www.gelcore.com). A large number of materials, substrates, epitaxy, packaging, and manufacturing companies have entered the LED market. In February 2006, lighting industry directory Lightsearch.com listed 71 companies that produced LED lamps.22 Most companies operate at a single stage of the supply chain, illus- trated in Figure 6. For example, companies that perform epitaxy do not usually do manufacturing or packaging. Likewise, most companies that focus on packaging do not produce raw materials or substrates. Moreover, companies that focus on basic R&D do not operate in the rest of semiconductor production.23 Even on a national level, specialization sometimes occurs. For instance, Taiwan is strong in R&D and manufacturing of LEDs, whereas Korea specializes in packaging, and China, a late entrant, is setting up epitaxy, and wafer and chip production 22â Recall that âlampâ as used in the traditional lighting industry means âbulb,â and note that the former term is most appropriate for LED lighting as no glass bulb is involved. Lamp here means a light-producing device, not a fixture. 23â Based on information on semiconductor companies in the LED industry gathered from solid- statelighting.net (2006).
182 INNOVATION IN GLOBAL INDUSTRIES (Wang and Shen, 2005). In addition, some countries specialize in the production of specific LED colors: Taiwan holds a majority market share for blue GaN LEDs at 34 percent, closely followed by Japan at 33 percent, whereas the United States and Korea lag with 19 and 12 percent, respectively (Wang and Shen, 2005). Although LEDs are still a small subset of the semiconductor market, they have rapid growth in demand, making this an attractive market for new and existing semiconductor firms. LEDs offer opportunities for semiconductor firms to diversify into a new market that promises long-term growth potential. For example, Avago Technologies, the worldâs largest privately held semiconductor company, recently announced three new series of HB full-color LEDs for the outdoor electronic signs and signals market (Business Wire, 2006). At the other end of the supply chain, LED âintegratorsâ like Color Kinetics play an important role in LED lighting. Since its establishment in 1997, Color Kinetics has built an impressive patented portfolio of these technologies, which it uses in LED lighting systems. Color Kinetics has pioneered intelligent LED systems that are networked and has created a new niche as a âsystems solutionsâ and lighting control technology provider. Color Kinetics has initiated several major projects that integrate LED lights with sound, movement, and rhythm through digital controls, and is working on white light systems. A subway tunnel in Chicago, for example, is bathed in several colors of LED light that periodically change (giving the impression of a sunset). The company leverages its strengths in innovation and engineering and works with selected Chinese manufacturers to assemble systems. In August 2007, Philips acquired Color Kinetics, which be- came Philips Solid-State Lighting Solutions, a business unit of Philips Lightingâs Luminaires group. LED LIGHTING R&D The big three producers are dominant in traditional lighting technology, as shown earlier using data on patents for these technologies. In this section we analyze the R&D positions of these and other firms for LED lighting. Methodology: Analysis of Patent Data To assess trends in the global location of LED lighting R&D we use patent data. Patent data yield information on successful R&D outputs. Although the information is partial because many inventions and innovations are not patented, within an industry patents are highly correlated with R&D spending and are in- dicative of R&D success. Moreover, patents yield relatively defensible property rights and hence represent an important component of the value of firmsâ R&D outputs. To analyze LED-related patents that pertain to lighting, a search criterion is needed to identify relevant patents. Choice of a criterion involves a trade-off
LIGHTING 183 between finding a subset of mainly relevant patents versus finding all relevant patents mixed with many more non-LED and non-lighting patents. We therefore chose a criterion to identify mainly relevant patents at the cost of excluding some LED lighting-related patents.24 All patents were identified whose title contains the key word or phrase âLED(s),â âOLED(s),â âL.E.D.(s),â or âlight emitting diode(s),â or the equivalents of the latter phrase in German, French, Spanish, Italian, or Portuguese.25 Other languages, including Asia-specific languages, ef- fectively are almost always included in our search because for almost all other languages the database we used has English translations of titles. The search criterion includes LED applications, including LED-type displays and (rarely) lasers, as well as LEDs whose glow is bright enough for general illumination. All patents granted are included regardless of whether they originated from firms, government programs, or university research labs. Since both the traditional and the LED lighting industries are global in terms of the firms involved and startup efforts, we obtained data for patents issued in most nations worldwide, although we focus initially on patents granted in the United States and Europe.26 For logistical reasons (we had to look up information by hand from actual copies of thousands of patents), we restricted the sample in the latter analysis to patents whose title included either âLED(s)â and âlighting,â or the phrase âlight emitting diode(s).â Our primary focus on U.S. and European patents addresses concerns that patents from other nations may face quite differ- ent approval requirements. Patents are counted only once if the identical patent is filed multiple times 24â International patent classification systems do not identify specific categories for LEDs or for LED lighting, and we draw on international data for which this limitation applies. 25â The non-English terms used are âlichtemittierende Diode(n),â âLeuchtdiode(n),â âdiode(s) luminescente(s),â âdiodo(s) electroluminoso(s),â âdiodo luminescente,â âdiodi luminescenti,â âdiodo(s) emissores de luz,â and âdiodo(s) emitindo-se claro(s).â The verb âledâ sometimes appears in titles for reasons unrelated to light emitting diodes, so we read all patent titles and eliminated patents for which âledâ was used as a verb. 26â Data are obtained from the Espacenet worldwide patent database, maintained by the European Patent Office (EPO). The data include patents granted by the relevant patent authorities in almost all nations worldwide (a detailed listing is available from the Espacenet website), including not only the most developed world but also Eastern European and developing Asian, Middle Eastern, and African nations (or cooperating regions) with significant innovative activity. The European patent authorities for which LED patents appear in the primary sample are the European Patent Office plus the national offices of Austria, Finland, France, Germany, Italy, Sweden, and the United Kingdom. Patents from former Soviet-bloc nations are excluded. In our secondary sample of patents from patent authorities worldwide, patent authorities that actually granted patents in the data are (ordered by continent) the United States, Canada, Mexico; Germany, Great Britain, Belgium, France, Netherlands, Hungary, former Soviet Union, European Patent Office; Japan, Korea, Taiwan, Singapore, India, China, Hong Kong; Australia, New Zealand; and South Africa.
184 INNOVATION IN GLOBAL INDUSTRIES in different nations.27 Patents granted in nations outside the United States and Europe are considered after our main analyses. Analysis of LED Patenting The analyses that follow compare the headquarters nationality of patenting firms and also the national R&D locations where invention was carried out. The headquarters location of a firm was identified as the international headquarters nation of the firm to which a patent was assigned. If a firm was owned by a âparentâ firm, we use the headquarters nation of the parent firm. Rarely, patents were applied for by multiple firms or individuals, and assignee credit was divided equally among these applicants. The R&D location where invention was carried out was determined by the nation listed in the address of each inventor named on the patents. Since inventorsâ addresses are not generally available in electronic bibliographic data, we looked up the nation for each inventor using the original patent documents. Rarely, different inventors of a single patent had addresses in different nations, in which case credit for each R&D nation was divided in proportion to the number of inventors in each nation. LED patent data are compared in two 4-year periods a decade apart, 1990- 1993 and 2000-2003. These 4-year periods ensure an adequate-sized, representa- tive sample of patent activity. Comparing between periods facilitates analysis of trends in R&D activity in LED lighting. As LEDs have developed growing markets in new applications, so has LED R&D grown. Based on U.S. and European patents, the number of LED patents granted quintupled from 200 in 1990-1993 to 1,000 in 2000-2003, as shown in Figure 7.28 Globalization of LED R&D: U.S. and European Patents The locations of inventors as reported on patent applications reflect where R&D occurred. We therefore assessed relative inventive activity in each nation, for 1990-1993 in Figure 8 and 2000-2003 in Figure 9, by determining the percent- 27â Equivalent patents filed in multiple nations are identified as catalogued on the EPOâs Espacenet patent web server, which defines equivalents based on identical priority claims. 28â Substantial growth also holds using our worldwide data for a narrower range of patent titles, with LED patents granted growing from 438 in 1990-1993 to 1,114 in 2000-2003.
LIGHTING 185 1000 800 2003 600 Patents 2002 400 2001 200 1993 1992 1991 2000 1990 0 1990s 2000s FIGURE 7â LED patents in sample granted during 1990-1993 and 2000-2003. SOURCE: lighting-7.eps Authorsâ analysis of patents granted in the United States and Europe (see text). Taiw an UK Netherlands Belgium 2% 2% 1% 1% Other France 2% Korea 2% 3% Sw eden 3% Germany US 8% 48% Japan 28% FIGURE 8â Invention locations of LED patents granted in 1990-1993. SOURCE: Au- thorsâ analysis of patents granted in the United States and Europe (see text). Lighting 8 NEW
186 INNOVATION IN GLOBAL INDUSTRIES Italy Korea China France Canada 2% 1% 1% 2% 2% Other 3% UK 2% Netherlands 2% Germany 11% US 46% Japan 13% Taiwan 15% FIGURE 9â Invention locations of LED patents granted in 2000-2003. SOURCE: Au- thorsâ analysis of patents granted in the United States and Europe (see text). Lighting 9 NEW ages of patents with inventors in each nation.29 In the period 1990-1993, inven- tors located in the United States predominated with 47 percent of all of the LED patents. Inventors in Japan held second place with 27 percent of LED patents. Inventors in Germany ranked third with 8 percent of LED patents, followed by inventors in Sweden, Korea, Taiwan, France, and the United Kingdom, each with 2-3 percent of LED patents.30 29â When computing the percentages of LED invention done in different nations, one concern is whether some nationsâ inventions might be of systematically poor quality, so that while those nationsâ inventors appear to accomplish a lot of R&D, in reality the value of their R&D output is much lower. One means to check whether this is the case is to examine only those relatively high-value patents for which firms went to the expense and trouble of obtaining equivalent patents in multiple countries or on multiple continents. Hence, we also examined patents using these criteria. These analyses yielded conclusions similar to the results reported in the text. 30â Percentages in this section often represent small numbers of patents, so readers are cautioned that small variations in patenting activity can sometimes lead to substantial changes in the precise percent- ages reported. Our primarily conclusions, however, are robust to typical random variations, with statis- tical significance tests reported in footnotes. Differences in arrival rates of patents in 1990-1993 were statistically significant between the United States and Japan (p = .0014), between Japan and Germany (p < .001), and (marginally) between Germany and Sweden (p = .0525), using exact p-values from exact Poisson regressions. (Here and below, exact Poisson tests are actually close approximations
LIGHTING 187 In the period 2000-2003, U.S. inventorsâ share of LED patents remained dominant at 47 percent. Rapid growth in inventions occurred in Taiwan, whose inventors achieved a 15 percent share of LED patents, ahead of Japanese inven- tors whose share had fallen to third place with 13 percent of LED patents. Ger- man inventorsâ share grew slightly to 11 percent of LED patents.31 Other nationsâ share of LED patent invention remained at 2 percent or less. Not only did LED innovation become somewhat more international, but non- U.S. companies became more international in the locations where they carried out research. Location of R&D (i.e., of inventors) is compared to locations of corpo- rate headquarters in Table 2 for 1990-1993 and Table 3 for 2000-2003. During the period 1990-1993, the United States was the only country whose companies supported substantial LED R&D abroad.32 LED patents were filed by inventors in Japan, the United Kingdom, and Germany for firms headquartered in the United States. Of the nine U.S. patents that had R&D locations abroad, Eastman Kodak was assignee for four, with inventors in Japan, and the remaining five had inven- tors in the United Kingdom (twice), France, Japan, and Malaysia. By 2000-2003, more companies were supporting R&D across the globe. Overall, companies in 13 countries sponsored LED R&D abroad. U.S. compa- nies, however, kept 95.1 percent of R&D within the United States. Asian LED invention sites for U.S. firms fell from 5.8 to 1.4 percent, and European sites for U.S. firms fell from 2.9 to 1.9 percent. In the period 2000-2003, U.S. companiesâ foreign-invented patents had inventors in Canada (1.6 percent) and nine other na- tions (each less than 1 percent). European companies began to support R&D in the United States, which now yielded 22.9 percent of European companiesâ LED patents. Asian companies also began to carry out R&D in the United States, yield- ing 3.6 percent of Asian companiesâ LED patents. Furthermore, Europe yielded 0.9 percent of Asian companiesâ LED patents.33 because they use rounded-to-integer values of weighted patent counts, rather than making assumptions about statistical (in)dependence between locations of inventors of some patents.) 31â The differences between Taiwanese versus Japanese, and Japanese versus German, patent inven- tion in 2000-2003 each involve 16-17 patents, and it is difficult to tell whether these reflect statistically meaningful differences in nationsâ propensities to generate LED inventions; these nationsâ rankings reported here should not be construed to indicate practically or statistically significant differences. (Exact significance tests are not feasible to compute in these cases, because the numbers of patents involved make exact Poisson calculations highly computationally intensive.) German inventorsâ arrival rate of patents was statistically significantly (p < .001) greater than the highest-ranked of nations with less invention, using an exact p-value from exact Poisson regression. 32â This difference between the United States and other nations is marginally statistically sig- nificant, p = .059, using Fisherâs exact test (two-tailed). (The test is approximate because it uses rounded-to-integer values of weighted patent counts, rather than making assumptions about statistical (in)dependence between locations of inventors of some patents.) 33â Use of inventors outside the home country in 2000-2003 was statistically significantly less for United States patent applicants than for non-United States firms (p < .001 using Fisherâs exact test as in the preceding footnote), with no significant difference between United States and Asian patent applicants.
188 INNOVATION IN GLOBAL INDUSTRIES TABLE 2â Location of Inventor vs. Firm Headquarters, 1990-1993 Location of R&D Location of HQ Asia Europe U.S. Other Asia 98.4% 1.6% Europe 100.0% U.S. 5.8% 2.9% 91.3% Other 100.0% SOURCE: Authorsâ analysis of patents granted in the United States and Europe (see text). TABLE 3â Location of Inventor vs. Firm Headquarters, 2000-2003 Location of R&D Location of HQ Asia Europe U.S. Other Asia 95.2% â 0.9% â 3.6% ââ 0.3% Europe â 0.7% 76.0% 22.9% ââ 0.4% U.S. â 1.4% â 1.9% 95.1% ââ 1.6% Other 100.0% SOURCE: Authorsâ analysis of patents granted in the United States and Europe (see text). To a slight degree, the United States became an innovation hub for compa- nies headquartered in other countries, mainly for Philips. Of the LED patents assigned to Dutch companies, 67.3 percent were filed by inventors located in the United States, as were only 1.8 percent of patents assigned to German companies. Germany also was an R&D source for 7.4 percent of Dutch firmsâ patents. These shifts largely reflect the internationalization of Philips as it acquired other firms and joint ventures. There was a corresponding shift in the number of companies using inventors in other countries. In the period 1990-1993, six U.S.-headquartered companies sponsored LED research abroad, totaling nine patents, while only one European and one Japanese company sponsored LED research abroad with only one such patent each. In the period 2000-2003, 25 companies located in all parts of the world sponsored LED research abroad, totaling 95 patents. This a substantial shift, but far from complete globalization, as even in 2000-2003 only 5.2 percent of LED patents involved work outside companiesâ headquarters nations; the vast majority of patents still used inventors in firmsâ home countries. As Table 4 shows, there was a shift from 1990-1993 to 2000-2003 in the dominant firms in LED patenting.34 Dominant firms are ranked here in LED ap- plications generally, including LED-type displays and LED backlights, not just 34â The shift in firmsâ rank ordering from 1990-1993 to 2000-2003 is statistically significant (p < .001) using Wilcoxonâs signed-rank test.
LIGHTING 189 TABLE 4â Leading Firms in LED Patenting, 2000-2003 versus 1990-1993 Rank Rank Headquarters Patents 2000-2003 1990-1993 Company Country 2000-2003 1 32 Philips/Lumiledsa Netherlands/U.S. 69.0 2 9 OSRAM/Siemensb Germany 41.0 3 â GE/Gelcorec U.S. 26.5 4 â United Epitaxy/Epistard Taiwan 17.2 5 3 Xerox U.S. 14.5 5 13 Oki Japan 14.5 7 1 Eastman Kodak U.S. 14.0 8 2 HPe U.S. 9.0 8 â Para Light Electronics Taiwan 9.0 10 â Hon Hai Precision Taiwan 8.0 11 â IBM U.S. 7.0 11 â Lite On Electronics Taiwan 7.0 13 8 Samsung Korea 6.8 14 â Nichia Japan 6.0 15 10 Cree U.S. 5.3 NOTE: Individuals and institutions were excluded from the table SOURCE: Authorsâ analysis of patents granted in the Untied States and Europe (see text). aIncludes all patents by Philips and its subsidiary Lumileds, Agilent as former joint venture partner of Lumileds, and Color Kinetics acquired in 2007. bIncludes all patents by OSRAM-Sylvania and its parent company Siemens, as well as 6.5 patents held by Patra Patent Treuhand. cIncludes all patents by GE and its subsidiary Gelcore. dIncludes all patents by United Epitaxy, Epistar with which it merged in 2005, and Epitech with which it merged in 2007. eHP patents in the 1990s pertain to technology such as LED print heads that may not have been owned by HPâs spin-off Agilent, which became part of Philips. LEDs for general illumination. The listed firms are unlikely to include materials makers because of the search criteria used. Only 4 of the top 10 firms that filed LED patents in 1990-1993 remained in the top 10 a decade later (OSRAM plus its parent firm Siemens, Xerox, Eastman Kodak, and HP) and 4 entirely new LED patent assignees had appeared in the new top 10. All of the big three traditional lighting firms featured at the top of LED patenting in 2000-2003 (Philips ranked first, OSRAM second, and GE third). Of these big three, Philips and (through Siemens) OSRAM had LED patents in 1990-1993. The emergence of the big three as the dominant LED firms in 2000-2003 can clearly be attributed to their joint ventures that allowed them to enter fully into the semiconductor-based LED market. Many LED patents came from firms not in the traditional lighting industry, including established semiconductor firms. Semiconductor firm United Epitaxy, founded in 1993, is one of several such Taiwanese firms. Firmsâ ranks are
190 INNOVATION IN GLOBAL INDUSTRIES measured somewhat noisily here, because the sample of patents used does not cover every LED patent (as noted in our earlier description of patent methods). 35 Nonetheless, the evidence shows an important role in LED technology of Asian firms, representing half the firms listed in Table 4.36 Hence, the big three lighting firms have managed to established dominant positions in LED lighting technology. Nonetheless, if LED technology develops as anticipated there may be greater participation by firms in Japan, Taiwan, and perhaps Korea and other Asian nations in the global lighting industry. Globalization of LED R&D: Worldwide Patents The results differ somewhat when patents from Japan, Taiwan, and other nations are included. These patents were initially excluded because of concerns over whether patents are of comparable quality and value in different nations and because the U.S. and European markets have been two of the worldâs larg- est. However, focusing only on U.S. and European patents may introduce a bias because some applicants develop R&D competence but apply for patents only in their home countries. Also there may be international differences in propensities to patent in different markets. Filings by individuals (instead of companies) show- case the differences; in the period 2000-2003 almost four times as many patents were granted worldwide as in the United States and Europe, and almost five times as many individuals were granted patent rights, using our worldwide sample of data with a restricted definition of LED patents. The majority of individual filings originated in Taiwan (35.6 percent) and Korea (28.6 percent). Among our sample of LED patent invention worldwide, in 1990-1993 Japa- nese inventors led with 78.3 percent of LED patents, the United States followed with 10.7 percent, and all other countries each invented less than 3 percent. A decade later in 2000-2003, Japanese invented 41.6 percent of LED patents, Taiwanese 22.0 percent, Americans 13.9 percent, and Koreans 12.9 percent of patents. Clearly by 2000-2003 more countries, notably Taiwan and Korea, were 35â Errors in ranking could result if certain companies are less likely than others to include our search terms in their patent titles, and our purpose is not to compute exact rankings. However, the rankings computed confirm industry impressions and are approximately valid absent any odd variations in firmsâ choice of terminology in patent titles: the difference between Philips and OSRAM in their ar- rival rate of patents was statistically significant (p = .010), the difference between OSRAM and GE was not statistically significant at conventional levels (p = .114), the difference between OSRAM and United Epitaxy/Epistar was statistically significant (p = .002), and the difference between GE and United Epitaxy/Epistar was not statistically significant (p = .174), using exact p-values from exact Poisson regressions. 36â cumulative measures of LED patenting were considered, Asia would likewise emerge as If playing an important role. Asian countries, particularly Japan, had strong early LED R&D, which is apparent in the 1990-1993 patent data.
LIGHTING 191 locations for LED R&D.37 Hence, these results indicate a possibly greater role of Asian inventors, when patents applied for in Asian nations are considered along with those in U.S. and European nations. Other Indicators of R&D Other indicators of globalization and Asian strength in LED innovation are international joint ventures and licensing agreements. Joint ventures in LEDs occurred between each of the big three traditional lighting firms and other in- ternational firms, as discussed earlier, all in 1999, with all three subsequently acquiring the joint ventures. International cross-licensing agreements, listed in Table 5, now exist between Philips (Dutch) and OSRAM (German), Philips and Nichia (Japanese), Philips and Toyoda Gosei (Japanese), OSRAM and Nichia, OSRAM and Seoul Semiconductor (Korean), OSRAM and Avago (American), Cree (American) and Nichia, and Cree and Seoul Semiconductor. The table also indicates other international licensing, disputes, and strategic research partner- ships. The evidence indicates considerable global dispersion and a growing Asian contribution in LED innovation. NATIONAL PROGRAMS AS INNOVATION DRIVERS Promoting R&D The development and market penetration of LEDs is closely linked with government policies and national programs. This is not uncommon in the semi- conductor industry. For example, Japan saw extensive growth in semiconductor R&D, which displaced U.S. leadership in the DRAM market, following a mid- 37â Japanese firms dominate the rankings when patents granted worldwide are considered. In the period 1990-1993, 9 of the top 10 ranked firms were Japanese. The five highest ranking firms by LED patents were Hitachi (35 patents), NEC (31), Toshiba (30), Mitsubishi (27), and Sanyo (23). Eastman Kodak (United States) was the only non-Japanese firm in the top 10 during the period 1990-1993. By 2000-2003, the top 5 ranking firms for LED patents were also Japanese: Nichia (38), Hitachi (30), Sharp (29), Showa Denko (26), and Citizen (25), and 8 of the top 10 were Japanese. Even among the Japanese firms, however, only Hitachi, Sharp, and Matsushita stayed in the top 10 ranking over the decade. The dominance of Japanese firms could, however, reflect differences in patent systems such as how often multiple claims are combined in one patent, coupled with the fact that firms are typically most likely to apply for patents in their home countries as well as possible international locations. The two non-Japanese firms in the top 10 during the period 2000-2003 were Taiwanese: Epistar (founded in 1996) held sixth place with 24.5 patents, and United Epitaxy (founded in 1993) held ninth place with 18.5 patents. The big three traditional lighting firms did not make the top 10 by this measure: Philips/Lumileds is ranked in 12th place, Osram plus Siemens in 15th place, and GE plus Gelcore in 36th place. Two Korean firms, LG and Samsung, were in the top 20. When only patents granted on multiple continents (a measure of high value) are considered, U.S. inventors are responsible for a slightly larger percentage of the sample than indicated here.
192 INNOVATION IN GLOBAL INDUSTRIES TABLE 5â Globalization of LED Patents: Licenses, Alliances, and Disputes Philips GE Seoul Semi- (Lumileds) OSRAM Lumination Cree Nichia conductor Philips (Lumileds) OSRAM Cross- license GE Lumination Cree Chip supply from Cree Nichia Cross- Cross-license Strategic Cross- license alliance license Seoul Semi- Cross-license Cross- Lawsuits conductor license Epistar Dispute Dispute (resolved) Toyoda Gosei Cross- Settlement Strategic license partnership Citizen Dispute White LED (resolved) license ROHM White LED License license Lite-On White LED License license Kingbright Lawsuit License Others White LED Licenses; Investmentc licenses; lawsuitb cross-license; disputea NOTES: Cross-licenses are agreements to share rights to large numbers of LED patents. A âlicenseâ indicates that a firm acquired the right to use a patent owned by a firm at the top of the table. White LED licenses pertain specifically to patents for white-light-producing LEDs. A âdisputeâ indicates a patent lawsuit brought by a firm at the top of the table against another firm. The âsettlementâ between Nichia and Toyoda Gosei involves patent infringement lawsuits that had been brought by each firm against the other. Strategic alliances and strategic partnerships are joint LED technology development efforts involving the two companies indicated. aOSRAM has white LED patents licensed to Harvatek, Vishay, Samsung SEM, Everlight, Ya Hsin, Lednium (Optek); a cross-license with Avago; and a resolved dispute against Dominant. bCree has LED patents licensed to Cotco, and Stanley and a lawsuit against BridgeLux. cNichia has an investment in and chip-license agreement with Opto Tech. SOURCE: Ledsmagazine.com (2007).
LIGHTING 193 1970s research program (Macher et al., 2000). There appears to be a correla- tion between countriesâ national research programs for LEDs and innovative activity in those countries. Key LED programs exist in the United States, Japan, Taiwan, and South Korea, precisely those countries that dominate LED patent- ing. China recently announced programs targeted toward LED innovation and high-technology industries in general. Judging by the impact of similar research programs in other nations and import of U.S. and Taiwanese talent, China may become an additional key player in the LED industry. While national programs collaborate extensively with universities and re- search labs, such institutions account for only about 4 percent of all LED patents, reflecting the limited funding available for commercializing their basic research. Nonetheless, university spin-offs have often created major companies such as Cree (with a market capitalization of $1.69 billion and revenue of $385 mil- lion in 2005). Universities and research institutions appear most innovative in Taiwan and Korea, accounting for about one-half and one-fifth, respectively, of all LED patents filed by universities and research institutions. 38 The remainder is split fairly evenly among the United States, the United Kingdom, Japan, and Belgium. Interestingly, China also features, filing 9 percent of all LED patents by universities and research institutions. With the exception of Belgium, each of these countries has a national program dedicated to development of LED light- ing, with goals to improve energy efficiency and gain market share in general illumination, as outlined in Table 6. Often the dedicated lighting program benefits from other supporting legisla- tion or programs. For example, the U.S. initiative to develop LEDs may be partly driven by programs such as Vision 2020, an industry-led program to develop a technology roadmap for lighting, initiated by the U.S. Department of Energy. The programâs goals are to develop standards for lighting quality; increase demand for high-quality lighting solutions; strengthen education and credentials of light- ing professionals; provide R&D incentives to accelerate market penetration of advanced lighting sources and ballast technologies for superior quality, efficiency, and cost-effectiveness; and develop intelligent lighting controls and flexible lu- minaries/system delivery platforms (DOE, 2000). Apart from aims to establish integrated energy-efficient lighting systems, the program has also launched the Energy Star voluntary labeling program designed by the Environmental Protec- tion Agency (EPA) and the National Appliance Energy Conservation Act that bans low-efficiency magnetic ballasts. Grants awarded by the DOE in 2006 totaled nearly $60 million, with a further $12 million provided by contractors 38â Research output at universities has often been measured by journal publication rather than patents, but it would be difficult to use a publication-based measure here without possible language-related biases (non-English speakers frequently publish in non-English journals not catalogued in available publications databases). Our measure of patents rather than publications may be more pertinent to applied than to basic research. The numbers are based on all patents (including national patents) during the period 2000-2003.
TABLE 6â Major National Research Programs Pertaining to LEDs 194 Country, Fundsa Program Phase & Objectives Funding Organizations USAb Energy Act of 2005: â¢ 007-2009: support research, 2 $50 mil/year Partnership DOE, industry, estimated $42.1 mil/year Next Generation Lighting development, demonstration, and authorized from 2007 universities & laboratories (2003-2013); includes Initiative (NGLI) commercial applications to 2009 anticipated extension of â¢ 010-2013 extension 2 Extended authorization funding to allocate $50 mil/ year from 2010-2013 SSL Project Portfolio â¢ ompleted projects 2003-2005: C Total: $70.9 mil Partnership DOE, industry, (current projects in NGLI) six key research areas: quantum universities & laboratories efficiency, longevity, stability and control, packaging, infrastructure, and cost reduction â¢ urrent projects through 2008: LED C and OLED Japan Light for the 21st century, â¢ 998-2002 (first phase): develop 1 Yen 6 bil ($52 mil) Japan R&D Center of Metals estimated $7.5 mil/year New Energy and Industrial GaN-based LED technology for (JRCM) (1998-2008) Technology Development lighting applications Organization (NEDO) â¢ evelop 13% market penetration by D 13 companies and 2010 (over 477 new patents already universities filed in one year) â¢ roduce 120 lm/W and 80% P efficiency by 2010 Ministry of Education, â¢ inancial year 2004: develop F Yen 500 mil ($4.6 mil) Several universities, more Culture, Science & medical equipment and therapeutic than 20 companies Technology techniques based on LEDs â¢ imilar amounts of funding expected S $4.6 mil/year next 4 years (2005-2008): establish 2005-2008 the Yamaguchi-Ube Medical Innovation Centre (YUMIC) â¢ hite HB-LEDs W
South Koreac Semiconductor Lighting â¢ 993-1996 (R&D by LG, Samsung, 1 KOPTI receives $20 KOPTIâs costs covered estimated $59.4 mil/year National Program & universities, and Korea Research mil/year in funding 73.1% by government, (2001-2008); excludes KOPTI (Korea Photonics Institute): reduce use of glass, 16.5% by Gwanju âCity of equipment value, and Technology Institute) phosphors, heavy metals Light,â 10.4% by industry half of LED Valley as â¢ 999-2000 (business phase, JVs, 1 KOPTI equipment âfiber-to-homeâ production runs): meet environmental value: $65 mil regulations July 2006 â¢ ave $20 bil on energy S â¢ 001 (activation phase, growth to 2 more than 340 companies): produce 80 lm/W white LED by 2008 LED Valley Project â¢ 005-2008: develop HB-LED 2 $100 mil (first phase) Gwanju âCity of Light,â â¢ 005-2008: second phase of a 2 $430 mil (second mix of national and local photonics industry project for phase HB LED & government and private- HB-LEDs fiber-to-home) sector investment â¢ eploy fiber-to-the-home networks D Taiwan Next Generation Lighting â¢ 004-2005 (40 lm/W): improve 2 2004-2005: NT$383 Consortium of 11 companies estimated $4.0 mil/year Project performance of white LEDs mil ($11.5 mil) (2002-2005) â¢ econd phase (60 lm/W): 100 lm/W S output in labs National Science Council â¢ irca 2002-2004: producing highly C NT$12 mil ($0.4 mil) NSC department of science efficient LEDs & engineering â¢ ed to 14 new patents and 20 new L manufacturing process technologies 195 continued
TABLE 6â Continued 196 Country, Fundsa Program Phase & Objectives Funding Organizations Chinad Semiconductor Lighting â¢ ive parks established in Shanghai, F Total investment: Xiamen: three companies and estimated $248.8 mil/year Project Xiamen, Dalian, Nanchang, Yuan 10 bil ($1.2 bil), government & cooperation (2005-2010); only Shenzhen: establish industrial parks allocated as follows: with Taiwan; Dalian: JV includes initial investment with up-, mid-, and down-stream Xiamen: $1.9 mil between companies and for five parks and the products; first phase likely to be (with focus on optoÂ science & technology group; 5-year plan; not directly 2005-2010: collaboration with electronics), Dalian: Shenzen: university, local comparable to other Taiwan and specialists from Taiwan $150 mil, Shenzhen: city government support & nationsâ figures as this and U.S.; anticipate $19 bil LED initial investment 3 200 companies includes manufacturing industry by 2010 bil Yuan ($375 mil); site investments total 20 bil Yuan ($2.5 bil) over 3-5 years (2005-2010) National Solid State Lighting â¢ 015 goals: savings from large-scale 2 2006-2010: $44 mil 15 research institutions & project as part of 11th 5-Year conversion to LED; 100 bil kW/h 2,500 companies Plan annually by 2015 â¢ 50 lm/W LED and capture 40% of 1 incandescent market â¢ educe environmental pollution R â¢ evelop strong industrial base D â¢ nternational cooperation if necessary I E.U. Sixth Framework program â¢ trengthen science & S $1.3 bil earmarked for estimated $16.3-32.5 technology base for international nanoÂtechnology (with mil/year (2002-2006) competitiveness IST section) est. assuming 5-10% dedicated toward LED aThe yearly fund flow was estimated as an annual mean of all funding programs over the entire time range of the programs. All figures in US$. bPresidential budget for fiscal year 2006 includes request of $11 mil for SSL. cKorea also has a national program for LCD and displays, from 2004-2008. Key players are LG and Samsung. No funding information. dChinaâs â863 Program,â or National High Technology Research & Development Program, includes development of OLEDs as a focus. SOURCES (in order of table): DOE (2005, 2006), Japan Research and Development Center of Metalsâ National Project (2000), Stevenson (2005), Compound Semiconductor (2004), Chiu (2004), Yahoo! News Australia & NZ (2004), Tang (2006), Ledsmagazine.com (2005c, 2005a), Steele (2006), European Com- mission Community Research (2002).
LIGHTING 197 (DOE, 2006). Some 65 percent of the DOE grants were awarded to firms, with the remainder split about equally between research laboratories and universities. Similarly, Japan has an LED association that promotes R&D and standard- ization in the LED industry. As well as aiming for energy-efficient lamps, the as- sociation has established a medical innovation center that conducts R&D on LED use in medical equipment and therapeutics. A 1979 Energy Conservation Law in Japan, updated in 1999, has been a key driver of energy conservation in factories, buildings, machinery, and equipment. Japan is the second-largest government supporter of R&D in general, after the United States, investing $90.3 billion in 1997. Of this budget, $6.8 billion was allocated toward national energy-related R&Dâ64 percent public sector and 36 percent private sector (Dooley, 1999). South Koreaâs lighting program is supported by a government-backed orga- nization, Korea Photonics Technology Institute, which aims to produce 80 lm/W white LEDs by 2008 and invests $20 million per year. Funding stems mainly from the government (73.1 percent), but also from industry (10.4 percent) and the âCity of Light,â Gwanju (16.5 percent). Gwanju is the center of the LED Valley project in Korea, aimed at penetrating LEDs into television backlighting by 2006, car lighting by 2008, and domestic lighting by 2010. Investment is significant at $100 million for the development of HB LEDs (plus $430 million partly for fiber-to-the-home) from 2005 through 2008. In addition, the Korean private sec- tor, namely Samsung and LG, are investing in LCDs and OLEDs, using Koreaâs LED infrastructure as a platform. Chaebols such as Samsung and LG are doing so through their business units and research labs, as well as a partial spin-off in the case of LG, in which it still has a 60 percent equity stake. But there have also been new startups for epiwafer foundries, substrates/GaAs ICs, and fiber optic componentsâmany set up by researchers from Samsung and LG or by university professors (Whitaker and Adams, 2002). Taiwan has had support from the National Science Council for LED research. Together with a consortium of 11 companies, Taiwan invested $11.5 million in LED research and development during the period 2003-2005. The second phase, to produce high-efficiency LEDs, is expected to receive $0.4 million in funding. The goal is to produce 100 lm/W output efficiency of LED bulbs in laboratories. In addition, Taiwan has a 6-year national initiative on nanotechnology worth $700 million, some of which is dedicated toward LEDs (Liu, 2003). China has budgeted $44 million to address solid-state lighting R&D needs as part of its 11th Five Year plan. The program will include 15 research institutions and university labs, and more than 2,500 companies involved in LED wafers, chips, packaging, and applications (Steele, 2006). The country expects to be the largest market for LEDs in the world, although it acknowledges a 6- to 20- year lag behind Japan, Europe, and the United States in LED device technology (Steele, 2006). The key driver behind the lighting project is energy savings. The goal is to penetrate 40 percent of the Chinese incandescent lighting market with 150 lm/W LEDs. The program was responsible for the establishment of five in-
198 INNOVATION IN GLOBAL INDUSTRIES dustrial parks in China during 2004 and 2005, backed by government, company, and university investment. The objective of the program is to save 30 percent of energy spent on lighting, the same as generated by the Three Gorges Project, in the next 15-20 years. An underlying national solid-state lighting project by the Ministry of Science and Technology aims to reduce environmental pollution and improve technology to develop a strong industrial base. Apart from its dedicated semiconductor lighting project, China is investing heavily in the semiconductor and advanced material industries in general. China is also focused on collaborat- ing internationally to develop its semiconductor industry, recruiting talent par- ticularly from Taiwan and the United States. One aspect that stands out among national LED programs is that Europe ap- pears to be lagging behind the United States and Asian countries. The European Unionâs Fifth Framework Program funds five research areas: nanotechnology, genomics and biotechnology, information technology, aeronautics and space, and food safety and health risk. The funding for the period 2002-2006 is $17.5 billion. Of this, $3.4 billion is assigned to the Information Society Technologies program, which includes research into semiconductor technologies and LEDs. The program funds research institutions, universities, and other organizations. The lack of specific initiatives for LED innovation may explain European coun- triesâ minor share of LED patents. Some European countries have more specific programs dedicated toward LEDs. In September 2006, for example, the German Ministry of Education and BASF inaugurated a new research lab, the Joint Innovation Lab (JIL) (BASF, 2006). The JIL is a cooperative effort between 20 BASF experts and industrial and academic partners researching new materials in organic electronics, con- centrated particularly on OLEDs for organic photovoltaics and appliances in the lighting market (OPAL). The German Ministry of Education and Research intends to invest around $800 million in the OPAL project. In addition, BASF spends over $1 billion on R&D each year. It is hoped that the projects will strengthen Germanyâs position in the emerging market of organic electronics and create the scientific and technological basis for initiating the production of OLED-based lighting (A to Z of Materials, 2006). In the newer technology of OLEDs, much of the work is concentrated in research institutions and academia, both domestically and abroad. To be com- mercially viable, OLED research requires substantial infusion of capital. Foreign industry, heavily funded by their governments, could develop an insurmountable lead in the technology, making it very difficult for U.S. manufacturers to com- pete, if the U.S. government does not provide comparable support. With appropri- ate support from government and industry, commercialization could occur in as little as 5 to 8 years (Tsao, 2002). A push is also being made to pursue good white LEDs, the âholy grailâ of LED lighting. Analysis of PIDA data compiled by DigiTimes shows that each of the aforementioned countries is investing in white LEDs. The United States
LIGHTING 199 is investing $50 million over 10 years, Korea $23.4 million over 5 years, Japan $10.7 million over 5 years, Taiwan $4.6 million over 3 years, China $3.3 million over 3 years, and Europe $1.0 million over 4 years (Wang and Shen, 2005). Demand Drivers To spur innovation indirectly, regulations and incentives for energy-saving technologies can enhance demand for new lighting technologies. In a study com- paring U.S. and Japanese lighting industry conservation measures, Akashi et al. (2003) found that conservation can be encouraged by regulation, incentives, and awareness campaigns. The U.S. Energy Policy Act of 1992 prohibited manufac- turing and import of lamps that do not meet efficiency standards and mandated that lamp lumen output, efficiency, and life be printed on packaging, making it easier for consumers to compare and select more energy-efficient products. Nev- ertheless, consumers still experience considerable confusion in choosing lighting, particularly for residential settings. In new construction, builders have generally installed basic lighting packages that lack energy efficiency and other quality improvements in favor of lower capital costs (rather than lower operating costs). Bridging the gap between available lighting technology and consumer knowledge is a significant challenge and one that in Japan is met jointly by government and industry initiatives. Future diffusion of LED lighting may reflect patterns now apparent for CFLs, which, although more efficient than incandescent and halogen lamps, have achieved low penetration in the U.S. market. Only about 2 percent of sockets nationwide, and 4 percent in California, now use CFLs. Flicker, color, upfront cost, and other drawbacks have contributed to their slow adoption, so that greater energy efficiency alone seems insufficient to penetrate much of the market, although the efforts of Wal-Mart to promote CFLs may result in a significant change in consumer behavior. IMPLICATIONS AND POLICY RECOMMENDATIONS This chapter has documented a shift taking place in the lighting industry. Traditional lamps are being replaced with CFLs. While the early traditional lighting industry was dominated by three big companiesâGE, Philips, and OS- RAMâas production of lamps became commoditized competitive pressures in lighting increased. Lower prices and margins shifted production of traditional lamps to Asia, especially China, the largest source of lamp imports in the United States. Improvements in lamp efficiency led to the development of fluorescents and other types of lamps, which successfully penetrated commercial and indus- trial markets and are poised to enter U.S. residential markets after years of delay among consumers who lacked awareness and were unwilling to spend money up front for savings later on.
200 INNOVATION IN GLOBAL INDUSTRIES A new lighting technology, LEDs, is leading to a shift in how we view lighting. LEDs have already penetrated end-use markets for automobile brake lights, signs and displays, backlighting, and traffic signals. Investments in the development of white LEDs are setting the stage for the use of LEDs as general illumination and threaten the traditional lighting industry and its three big players. LEDs are a disruptive technology that has allowed many new players to enter the lighting market. While Japan and the United States dominate the LED market in terms of R&D and revenue, their market share is being eroded by fast-growing entrants, especially from Taiwan. Taiwan leads global production of blue (GaN) LEDs (Wang and Shen, 2005), had the second- or third-largest amount of LED patents by our counts in 2000-2003, and has two firms high on our LED patent ranking tables. Philips, OSRAM, and GE were not involved in the early stages of LED tech- nology development. It was only in 1999 that the big three decided to enter the LED market through a series of joint ventures that the companies later acquired. These firms may further build their strengths in this technology through acquisi- tion. In mid-June 2007, Philips acquired Color Kinetics for approximately $791 million.39 These big three firms appear to have established dominant positions in LED technology, judging from our patent analyses. However, it remains to be seen whether the big three will replicate the tight oligopoly they held in the traditional lighting industry in most of the 20th century. Partly this is because the semiconductor supply chain is fragmented as firms in this sector are typically not vertically integrated; by specializing, companies are able to keep costs down. Our analysis indicates that LED producers likewise operate at various stages of the supply chain and do not integrate vertically. This means that the LED market has witnessed many new entrants and has also created opportunities for new ventures in areas such as system controls and integration. Although LEDs have some clear advantages over traditional lamps, such as added flexibility, integration with digital systems, and higher energy savings, they are also still costly to produce. The question remains when (indeed whether) white LEDs will successfully displace traditional general illumination technolo- gies, especially among residential buyers. Evidence from CFL, HID, and other efficient traditional technologies shows low penetration rates among consumers. To aid success of LED lighting, therefore, governments might not only fund basic R&D but also promote awareness among consumers so that LED lighting products diffuse in the residential market. Governments worldwide are making significant investments into LED R&D and promotion of the technology. Govern- ment programs, such as the one in the United States, have allowed small startups 39â At the time of the announcement Color Kinetics had 71 patents granted and over 15,000 instal- lations. The merged entity will operate under the name Philips Solid-State Lighting Solutions, with intelligent and premium LED product lines ultimately co-branded Philips/Color Kinetics (Color Kinetics, 2007).
LIGHTING 201 and university research labs to make progress on LED R&D and gain a foothold in this new market. U.S. and Asian government programs, in particular, have made the largest investments. China, which is still at the early stages of ramping up capacity and technology to produce LEDs and therefore lagging behind other countries, is addressing R&D in solid-state lighting as part of its 11th Five Year Plan and is setting up five business parks dedicated to these new lighting technologies. China has a strong interest to meet its own energy-efficiency needs. Already, there is a trade imbalance between China and the United States for semiconductors gener- ally. In 2002, the United States imported $6.4 billion worth of semiconductor products from China, while exporting only $2.2 billion worth (Holtz-Eakin, 2005). Given its investments in R&D, China might become an important player in the global LED market. Analysis of these trends indicates that Asian countries such as Japan and Tai- wan, and possibly China and Korea, are poised to take an increased role in R&D, production, and diffusion of LED technology. Evidence provided by the patent analysis suggests a potential shift toward these Asian nations. Extensive public and private investment will help if the United States is to keep up with the op- portunities presented by these new technologies. Moreover, efforts to encourage consumers to use solid-state lighting as it becomes efficacious in new applications may help domestic markets to grow and support the commercialization of these important energy-saving technologies. ACKNOWLEDGMENTS Partial funding is gratefully acknowledged from the Alfred P. Sloan Founda- tion and from the Center for Future Energy Systems at Rensselaer Polytechnic Institute. Insightful comments on an earlier draft were kindly provided by Kevin Dowling, Jeffrey Macher, Bill McNeill, Klaus Minich, David Mowery, and David Simons. We also thank our discussants and anonymous referees at each stage of development of this work. REFERENCES Akashi, Y., L. Russell, M. Novello, and Y. Nakamura. (2003). Comparing Lighting Energy Con- servation Measures in the United States and Japan. Working Paper, Lighting Research Center, Rensselaer Polytechnic Institute. Arensman, R. (2005). LED market lights up. Electronic Business 31(4):24. Arthur, A. (2006). A unique history of the light bulb. Available at http://www.contentmart.com/ ContentMart/content.asp?LinkID=19298&CatID=328&content=1. Accessed December 2006. Ashdown, B. J., D. J. Bjornstad, G. Boudreau, M. V. Lapsa, B. Shumpert, and F. Southworth. (2004). Assessing Consumer Values and Supply-Chain Relationships for Solid-State Lighting Tech- nologies. Report ORNL/TM-2004/80, Oak Ridge National Laboratory, Oak Ridge, Tenn.
202 INNOVATION IN GLOBAL INDUSTRIES A to Z of Materials. (2006). BASF Open Organic Electronics Research Laboratory, 11 September. Available at http://www.azom.com/details.asp?newsID=6627. Accessed September 22, 2006. Barbaro, M. (2007). Power-sipping bulbs get backing from Wal-Mart. New York Times, January 2. BASF. (2006). Joint Innovation LabâOrganic Electronics: Research with customers and academic partners. Press release, September 11. Bellis, M. (2006a). The history of fluorescent lights. Available at http://inventors.about.com/library/ inventors/bl_fluorescent.htm. Accessed December 2006. Bellis, M. (2006b). Timeline of electrical lighting. Available at http://inventors.about.com/od/ lstartinventions/a/lighting_2.htm. Accessed December 2006. Bellis, M. (2006c). The history of the incandescent lightbulb. Available at http://inventors.about. com/library/inventors/bllight2.htm. Accessed December 2006. Bishop, A. L., R. Worrall, L. J. Spohr, H. J. McKenzie, and I. M. Barchia. (2004). Response of Culicoides spp. (Diptera: Ceratopogonidae) to light-emitting diodes. Australian Journal of Entomology 43(2):184-188. Bowers, B. (1982). A History of Electric Light & Power. Peter Pergrinus. Bright, J. R. (1958). Automation and Management. Boston, MA: Harvard University. Burgess, D. S. (2006). Efficient white OLED employs down-conversion. LED Focus, May 2006. Available at http://www.photonics.com/content/spectra/2006/May/LED/82586.aspx. Accessed December 2006. Business Wire. (2006). Avago Technologies introduces new high performance extra bright oval LEDs for outdoor electronic sign and signal applications. September 19. Calwell, C., C. Granda, L. Gordon, and M. Ton. (1999). Lighting the Way to Energy Savings: How Can We Transform Residential Lighting Markets? Volume 1: Strategies and Recommenda- tions. Prepared by Ecos Consulting for the Natural Resources Defense Council, San Francisco, Calif. Chinesewings. (2005). Philips Celebrate the Shanghai Innovation and Technology Park Lighting Project. August 1. Chiu, Y. T. (2004). Optical devices to replace toxic lighting mechanisms. Taipei Times. April 16. Color Kinetics. (2007). Color Kinetics Announces Acquisition by Philips. Press release, June 19. Compound Semiconductor. (March 2004). Taiwanese suppliers target white LED improvements. Compound Semiconductor. Compoundsemiconductor.net. (2004). Japanese LED project targets medical uses. Available at http:// compoundsemiconductor.net/articles/news/8/7/26/1. Cox, J. B. (2003). Semiconductor-maker Cree gets boost from Taiwan court ruling on patents. Knight Ridder Tribune Business News. September 26, p. 2. DOE (U.S. Department of Energy). (1993). Housing Characteristics 1993. Energy Information Ad- ministration, Office of Energy Markets and End Use, Report DOE/EIA-457. DOE. (1995). Commercial Buildings Energy Consumption Survey. Energy Information Administra- tion, Office of Energy Markets and End Use, Report DOE/EIA-871. DOE. (1997). Commercial Buildings Characteristics 1995. Energy Information Administration, Office of Energy Markets and End Use, Report DOE/EIS-E-0109. DOE. (2000). Vision 2020, The Lighting Technology Roadmap. A 20-Year Industry Plan for Lighting Technology. Office of Building Technology, State and Community Programs, Energy Efficiency and Renewable Energy. DOE. (2005). Energy Act Authorizes Next Generation Lighting Initiative. Available at http://www. netl.doe.gov/ssl/083105.html. Accessed December, 2006. DOE. (2006). 2006 Project Portfolio: Solid State Lighting. D&R International Ltd. for Building Technologies Program, Lighting Research and Development, Office of Energy Efficiency and Renewable Energy.
LIGHTING 203 Dooley, J. J. (1999). Energy R&D in Japan. Prepared for U.S. Department of Energy, contract DE- AC06-76RLO 1830. Pacific Northwest Laboratory, Battelle Memorial Institute. Document no. PNNL-12214. Economist. (2007a). An environmentally friendly bulb that may never need changing. The Economist June 19. Economist. (2007b). Everlasting light. The Economist, Technology Quarterly insert, September 6. Eisenberg, A. (2007). The television screen, sliced ever thinner. New York Times. December 23. European Commission Community Research. (December 2002). The Sixth Framework Programme in Brief. Brochure available at http://ec.europa.eu/research/fp6/pdf/fp6-in-brief_en.pdf. Accessed December 2006. Gartner. (2007). Gartnerâs Final Semiconductor Market Share Results Show Industry Grew 10 Percent in 2006. Available at http://www.gartner.com/it/page.jsp?id=503221. Accessed January 2008. General Electric. (2006). Lighting: The last 125 years. Available at http://www.gelighting.com/na/ business_lighting/education_resources/learn_about_light/history_of_light/last_years.htm. Ac- cessed December 2006. General Electric. (2007). Consumer & Industrial Announces Intention to Restructure Its Lighting Business to Capture Another Century of Growth and Leadership. Press release, October 4. Available at http://www.genewscenter.com/Content/Detail.asp?ReleaseID=2704&NewsAreaI D=2. Accessed November 1, 2007. Genlyte Thomas. (2005). Annual Report, 2005. Gizmodo. (2008). Samsungâs 31-inch OLED is biggest, thinnest yet. Available at http://gizmodo. com/342912/samsungs-31+inch-oled-is-biggest-thinnest-yet. Accessed February 20, 2008. Griffiths, T. (2006). Solid state lighting success is niche to niche. LIGHTTimes Online, December 11, p. 1. Hadley, S. W., J. M. MacDonald, M. Alley, J. Tomlinson, M. Simpson, and W. Miller. (2004). Emerging Energy-Efficient Technology in Buildings: Technology Characterizations for Energy Modelling. Prepared for the National Commission on Energy Policy. Oak Ridge National Labo- ratory, U.S. Department of Energy. Herkelrath, M., A. Laksberg, and L. Woods. (2005). A Brighter Future: Advances in LED Energy Efficient Lighting Technology. Manuscript, University of Washington. Holtz-Eakin, D. (2005). Economic Relationships Between the United States and China. Congressional Budget Office Testimony, Statement before the Committee on Ways and Means, U.S. House of Representatives, April 14. Hong, E., L. A. Conroy, and M. J. Scholand. (2005). U.S. Lighting Market Characterization, Vol- ume II: Energy Efficient Lighting Technology Options. Technical report prepared for Building Technologies Program, Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy by Navigant Consulting. Howard, W. E. (2004). Better displays with organic films. Scientific American 290(2):76-81. Inquirer. (2004). Inventor of LED honoured in his own day, strike a light. April 27. Japan Research and Development Center of Metalsâ National Project. (2000). Light for the 21st Century: The Development of Compound Semiconductors for High Efficiency Optoelectronic Conversion, Year 2000 Report of Results. (English translation, 2002). Ledsmagazine.com. (2005a). China promotes benefits of solid-state lighting. Available at http://leds- magazine.com/features/2/2/5/1. Accessed June 2005. Ledsmagazine.com. (2005b). LED industry in Taiwan set to grow by 16 percent. Available at http:// ledsmagazine.com/news/2/3/36/1. Accessed April 2007. Ledsmagazine.com. (2005c). Shenzhen starts to build LED manufacturing base. Available at http:// ledsmagazine.com/news/2/6/22/1. Accessed April 2007. Ledsmagazine.com. (2006). OLED focus for Osram Opto, efficacy reaches 64 lm/W. Available at http://www.ledsmagazine.com/news/3/8/8. Accessed February 2008.
204 INNOVATION IN GLOBAL INDUSTRIES Ledsmagazine.com. (2007). Seoul and Osram sign LED patent cross-license deal with Osram Opto Semiconductors. Available at http://ledsmagazine.com/news/4/8/33. Accessed October 2007. Leonard, R. S. (1992). Lighting the path to profit: GEâs control of the electric lamp industry, 1892- 1941. Business History Review 66(2):305-335. LIGHTimes. (2006). Wal-Mart sees cold cash in LED holiday lighting and LED lighting for refrigera- tor cases. November 21, p.1. LIGHTimes. (2007). Quiet big news announces cross licensing of all Osram and Philips inorganic and organic LEDs. Available at http://www.solidstatelighting.net/lightimes/?date=2007-02-02. Accessed February 2, 2007. Lightsearch Directory. www.lightsearch.com. Accessed June 2006. Liu, Y. S. (2003). III-V Compound Semiconductor Industry and Technology Development in Tai- wan. International Conference on Compound Semiconductor Manufacturing. GaAsMANTECH, Inc. Maccagno, P. (2002). Overview of the high brightness LED market. In Light Emitting Diodes 2002: The Strategic Summit for LEDs in Illumination, conference proceedings, San Diego, Calif., October. Macher, J. T., D. C. Mowery, and D. A. Hodges. (2000). Semiconductors. Pp. 245-285 in U.S. Industry in 2000: Studies in Competitive Performance, D. C. Mowery, ed. Washington, D.C.: National Academy Press. Mintel. 2003. LightbulbsâU.S.âMay 2003. Mintel International Group. Monopolies Commission. (1968). Second Report on the Supply of Electric Lamps. London: Her Majestyâs Stationery Office. Navigant Consulting. (2003a). Energy Savings Potential of Solid State Lighting in General Illumi- nation Applications. Technical report prepared for Building Technologies Program, Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy. Navigant Consulting. (2003b). Energy Savings Estimates of Light Emitting Diodes in Niche Lighting Applications. Technical report prepared for Building Technologies Program, Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy. OLLA Project Report. (2006a). Demonstrate a white p-i-n type OLED. High brightness OLEDs for ICT & next generation lighting applications funded under the IST priority (contract nr 4607) of the European 6th Framework Programme. February 2006. OLLA Project Report. (2006b). White pin-OLED with improved polymer injection layer and effi- ciency above 10 lm/W. High brightness OLEDs for ICT & next generation lighting applications funded under the IST priority (contract nr 4607) of the European 6th Framework Programme. August 2006. OSRAM. (2006). Highlights from 100 years of the OSRAM brand. Available at http://www.osram. com/cgi-bin/press/archiv.pl?id=450. Accessed December 2006. OSRAM China Lighting Ltd. Available at http://www.osram.com.cn/aboutus/index.jsp (in Chinese). Accessed December 2006. Philips. (2006). Annual Report 2005. Philips. (2006). A history of Philips and Lighting. A century of innovation at Philips Lighting. Avail- able at http://www.lighting.philips.com/gb_en/about/sub_feature_4. php?main=gb_en&parent =1&id=gb_en_about&lang=en. Accessed December 2006. Physorg.com. (2005). Universal Display First to Achieve 30 Lumens Per Watt White OLED. Au- gust 5. RPI (Rensselaer Polytechnic Institute). (2006). Rensselaer Researchers Aim to Close âGreen Gapâ in LED Technology. Press release, August 23. Sandahl, L. J., T. L. Gilbride, M. R. Ledbetter, H. E. Steward, and C. Calwell. (2006). Compact Fluorescent Lighting in America: Lessons Learned on the Way to Market. Research Report Prepared for the U.S. Department of Energy under Contract DE-AC05-76RLO 1830. Pacific Northwest National Laboratory, Richland, Wash.
LIGHTING 205 Siemens. (2006). Lighting. Available at http://w4.siemens.de/archiv/en/innovationen/licht.html. Ac- cessed December 2006. Solidstatelighting.net. (2006). Directories. Available at http://www.sslighting.net/cts/. Accessed June 2006. Steele, B. (2006). China pours millions into solid-state lighting program. Ledsmagazine.com, July 25. Available at http://ledsmagazine.com/news/3/7/21/1. Accessed December 2006. Stevenson, R. (2005). Oil-free Korea prioritizes solid-state lighting project. Available at http:// compoundsemiconductor.net/articles/magazine/11/4/4/1. Accessed December 2006. Taiwan Economic News. (2004a). Taiwan poised to retain no. 2 place in world LED production. March 24. Taiwan Economic News. (2004b). Taiwanese led suppliers threatened by Japanese rivalsâ bold ex- pansion plans. April 29. Tang, Y. K. (2006). Lighten up: Chinaâs resolve to embrace energy-efficient lighting unleashes a powerful market potential. Beijing Review, March 2, updated December 13, 2006. Available at http://www.bjreview.com.cn/science/txt/2006-12/13/content_50654_2.htm. Toniolo, T. (2006a). Shining a light on quality. Product Design & Development 61(11):14-15. Toniolo, T. (2006b). What determines an LEDâs quality? Electronic Products (June). Toshiba. (2006). The History of Toshiba lighting. Available at http://www.tlt.co.jp/tlt/english/com- pany/company.htm and http://www.toshiba.co.jp/worldwide/about/ history.html#1875. Ac- cessed December 2006. Tsao, J. Y., ed. (2002). Light Emitting Diodes (LEDs) for General Illumination: An OIDA Technology Roadmap Update 2002. Optoelectronics Industry Development Association. Walker, R. C. (2004). Cutting through the buzz production of HB-LEDs in Taiwan, South Korea and China. CompoundSemiNews. Available at: http://www.compoundsemi.com/documnts/article/ news/3850.html. Accessed May 2006. Wang, M., and J. Shen. (2005). LEDs lighting the future: Q&A with Biing-jye Lee, president of Epistar. Ledsmagazine.com, November. Whitaker, T., and B. Adams. (2002). Korea is the new compound semiconductor boom region. Com- poundsemiconductor.net, April. Wikipedia. (2006). Light-emitting diode. Available at http://www.wikipedia.org/wiki/LED. Accessed August 30, 2006. Williams, B. (2005). A history of light and lighting. Edition: 2.3. Available at http://www.mts. net/~william5/history/hol.htm. Accessed December 2006. Yahoo! News Australia and New Zealand. (2004). China launches 50 semiconductor lighting proj- ects. Available at http://au.news.yahoo.com/040722/3/q08l.html, July 22. Accessed December 2006. Zou, H. (2002). Envoy of brightness. Shanghai Star. January 24.