'Past is prologue' for emerging illumination technologies

The last 100 years of lighting-market history offers useful insights for newcomers, according to DARPA program manager Douglas Kirkpatrick. In an interview with Senior Editor Hassaun A. Jones-Bey, he explains.

LFW: Could you summarize the most relevant historical trends or factors?

Douglas Kirkpatrick: From the U.S. Department of Defense (DOD) perspective, there are two key questions: What is a realistic timeline for initial DOD and commercial SSL product availability? And what is a realistic timeline for the broad commercial availability of the technology? The first step is to look at the major technology-marketplace shifts that have occurred over the past 100 years and look at how those technologies went from the lab to DOD or commercial application. What elements of leverage did the new technology products use to displace the legacy producs? How long did this process take? What were the entry conditions and what were the timescale and scope of eventual saturation?

The elements of leverage for lighting can be reduced to two primary categories: performance and cost. Cost includes all of those elements with a financial impact: first cost, efficiency, socket and fixture compatibility, lifetime, maintenance, and possibly disposal cost. Performance includes such characteristics as color (correlated color temperature), color quality (spectral power distribution, percent-envelope from daylight, CRI), turn-on time, restrike time, and optical extent (point source or broad area source). The other important observation is that the lighting market is not the lighting market, but is an ensemble of many smaller markets, each with different prioritizations of the performance and cost driver elements.

Two observations scream at us from the historical data. The first is that while performance characteristics of a new technology determine what market segments it can compete in, cost characteristics determine the winners and losers. Note that the cost drivers can vary drastically between market segments—this is particularly true for the DOD market segment where first cost may be less important than maintenance or logistics costs. The second observation is that the time from successful market introduction to the beginning of the broader market saturation is 15 to 25 years. So if the first significant and successful SSL products don't hit the market until 2020, then significant market penetration probably is on a timeline for 2040, instead of current SSL market and performance projections of 2030.

LFW: How should generally accepted SSL market projections and planning change in light of this historical analysis?

DK: The early and effective product introduction of SSL technology is an essential component of any eventual success in the broader lighting market, and should be at least as high in priority as the longer term technology development efforts. The timeline of success places a premium on carefully selected early product introductions.

LFW: Could you give a specific example?

DK: Taking the DOD market segment as a specific example, early entry applications such as battery-efficient flashlights or laptop computer backlights have a high value, and can lead to corresponding early-entry commercial lighting niche markets. Follow-on applications such as remote source lighting (delivery to point-of-use via fiberoptics) could have significant benefits for the complete range of military platforms.

LFW: Could you describe DARPA's approach to advanced lighting?

DK: The DARPA program High Efficiency Distributed Lighting (HEDLight) is aimed at allowing DOD platform designers to optimize the layout of the electrical distribution network largely independent of the platform lighting requirements—by developing high-efficiency remote-source fiberoptic-distributed lighting systems. The program focuses on high total-system efficiency—from wall plug to light delivered to the task space—and this includes electrical efficiency and mass efficiency.

As with most DARPA programs, the HEDLight program is objective-focused and technology agnostic. Both HID and LED light sources are under development, and both have advantages. The HID development effort has made remarkable progress and is poised—with a little luck—to make some truly significant breakthroughs in the next year or two. The LED effort is also making phenomenal progress—but it is important to note that so-called "conventional" lighting technology is not standing still.

A second HEDLight objective is a sunlight-quality spectrum that avoids significant gaps in output spectrum. DOD systems go in harm's way and it is critical to be able to rapidly recognize and categorize damage. Incorrectly matching a tie to your suit is an inconvenience; incorrectly identifying a fluid leak on an aircraft can be fatal. So we weren't comfortable with the color-rendering index (CRI), a metric of spectral-power distribution at 8 or 14 selected wavelengths, because it has the potential to ignore a 20- to 30-nm gap in spectrum. Consequently, we devised a percent-deviation-from-daylight spectral-power-distribution metric to identify gaps in output spectrum as small as 10 nm. The most robust baseline for a target spectrum was the CIE definition of the daylight spectrum for the corresponding correlated color temperature (CCT). Such an output spectrum would also hopefully ease the visual transition for the sailor moving from the open deck of a warship to the sealed compartments below decks.

LFW: In suggesting a need to revise some SSL market perspectives and planning, you've also suggested that specific time windows must be met.

DK: The HEDLight program is targeting demonstration applications in the DOD in the 2006/07 timeframe, and larger scale DOD applications might start in earnest by 2009. Why the delay? The "owners" and developers of DOD systems need to see, characterize, verify, and validate the performance and operating advantages of the new technology product before they will risk incorporating it into their responsibility. An evaluation period of two years and a design-in period of an additional one to two years are very aggressive, and will only catch a few systems in the proper portion of their development or refit cycle. The bulk of the opportunity will lie beyond 2010.

SSL light sources are a part of the HEDLight program, and the pace of the productization path is probably representative for most commercial market segments.

This timeline suggests that significant market penetration by 2030 requires market entry by 2010. Some portion of that time has to be allocated to product engineering, pilot-scale implementation, ramp to full-scale manufacturing, and development of the appropriate sales channels. The base SSL technology development that feeds into potential products has to be completed before that process can begin, and to focus on the most relevant technology development efforts means that the list of potential products, the corresponding technology development requirements, and a reasoned estimate of the success probability of different development efforts—is needed now.

LFW: Is there else anything that you would care to add or emphasize?

DK: The pace of development in SSL has been impressive. When the HEDLight program was conceived 18 months ago, the possibility of SSL light sources was included mostly as a potential hedge investment. Since then, SSL capabilities have improved to where it is a competitor with more conventional approaches.