SIL program spans LED technology to the breadth of emerging SSL applications (MAGAZINE)
Quality of light was the predominant theme at Strategies in Light, reports MAURY WRIGHT, but the 2014 conference reflected a maturation of LED technology with speakers addressing a broad scope of topics, including new ways to approach the lighting market and surprising new applications for LEDs.
"How to make chicks happy" was one of the more compelling presentations at the 15th-anniversary Strategies in Light Conference (SIL) held February 25–27, 2014. LED lighting is how, the birds in question are layer or broiler chickens, and that session was representative of the broad range of applications presented at SIL in which solid-state lighting (SSL) is being applied. The 2014 SIL audience also heard plenty on the latest in LED components. And if there was a prevailing theme beyond diversity of applications, it was that the general lighting market is entering a stage where manufacturers must boost the value and quality of LED lighting products to both satisfy customers and to maintain economic health in the lighting industry.
Along with a packed exhibit floor (Fig. 1) SIL 2014 comprised a Plenary session, as well as market, technology, and luminaire design tracks. Moreover, workshops along with an Investor Forum that pitted early-stage companies against investors in the Dragon's Lair filled a pre-conference day. Here we will focus on the Plenary keynotes, along with some of the more interesting application-centric presentations.
Challenges to LED makers
The first keynote during the Plenary session was entitled "The real challenges facing the LED market" and came from Bangwon Oh, senior vice president of lighting marketing at Samsung Electronics. Oh was focused primarily on the replacement lamp market with the assumption that mid-power LEDs are the most likely path forward, but his message applies across LED segments because he called for an all-new LED structure.
Oh described the three primary challenges that he says face LED manufacturers: lower-cost components, better efficacy, and better color consistency. He addressed the issues one at a time starting with cost. The transition to lower-cost mid-power packages has provided tremendous benefits to the lamp makers, according to Oh. Moreover, he said that mid-power LEDs capable of operating at 150–200 mA, in what was previously considered the lower end of the high-power region, have enabled a 50% reduction in lamp prices and a 230% increase in lm/$ (lumens per dollar).
The ability to push typical mid-power packages, however, is diminishing. "There seems to be very little room for further cost reduction in mid-power platforms," said Oh. "Therefore, a total new platform is necessary to meet the market requirement for over 2000 lm/$ in a bulb."
Indeed, the existing mid-power architecture is also incapable of meeting better efficacy and color-quality goals, according to Oh. He said mid-power LEDs can only meet a 3-step MacAdam ellipse target through binning, whereas high-power LEDs can increasingly hit that target for the bulk of production. He added, "Improving chip structure is the most straightforward way."
Mid-power flip-chip
Oh said that Samsung is working on a flip-chip architecture for mid-power LEDs. Such an approach, where the epitaxial layers are grown on a patterned substrate with contacts constructed on the top of the epitaxial layers, has been used by a number of manufacturers in the high-power space. The chip is then flipped with the contacts available on the bottom of the structure and the light is extracted through the sapphire layer for maximum efficacy.
Bringing flip chips to mid-power will also enable changes in phosphor application that will yield better color consistency, according to Oh. Because there are no bond wires on the surface of the LED, unlike traditional mid-power structures, Samsung will apply phosphor in a film. Oh said the approach will lower cost and improve consistency (Fig. 2).
Oh called the new Samsung structure White Flip-Chip. The adjective "white" is simply the color of the sides of the flipped die that are not encapsulated in a plastic package. The approach is similar to the high-power chip-scale package that Philips Lumileds announced at SIL in 2013. Oh said the new structure would push mid-power LEDs above 1W and the cost profile near 1500 lm/$. Moreover, the design will maintain high efficacy at the higher drive currents. As we went to press, Samsung formally introduced the LED hinted at during the keynote.
Bringing the message back to the target lamp market, Oh offered some comparisons with current mid-power LEDs. He said a 60W-equivalent lamp that today requires 20 0.3W LEDs would only need 10 of the new devices. Moreover, he said the approach will yield smaller LEDs that offer better optical control for directional applications such as downlights.
SSL system benefits
Following Oh, Mark Rea, director of the Lighting Research Center (LRC) at Rensselaer Polytechnic Institute, took the stage to discuss SSL systems and how to monetize LED lighting technology (Fig. 3). We covered some of this ground in our interview with Rea leading up to SIL. But Rea was far more specific in his keynote.
An analogy led Rea into the topic at hand. He said that Audi-branded vehicles comprise about 20% of the total sales of The Volkswagen Group. Yet Audi sales contribute 40% of the profit to the group. "Audi is perceived as delivering greater value because it delivers more engineered benefits and that I suggest can be monetized as demonstrated by The Volkswagen Group," said Rea.
LEDs have breathed new life into lighting, leaving space for engineered benefits. Rea said that for the first time in his near 40-year career focused on lighting his neighbors are interested in the topic. But he warned that if the industry can't engineer tangible benefits, "It's close to a race to the bottom. It's going to be a price war again."
Benefit metrics
The problem with how the SSL industry is progressing is a focus on lumens because we can measure that data. But we don't see lumens and metrics such as lm/W and lm/$ are only meaningful in a commodity way. What people will pay a premium for, according to Rea, are things such as health and wellbeing, driver safety, personal security, and improved productivity. But he admitted those things are difficult to quantify.
One way to characterize lighting performance, however, is benefit metrics based on spectral weighting functions. The lighting and color science industry has developed numbers of characterizations of light and color that can be related specifically to benefits such as visual performance relative to work or driving a car; brightness relative to personal security; or what he calls Class A color relative to food, furnishings, and faces.
Indeed, if the industry established benefit metrics tied to specific applications, then you could relate those metrics to dollars or power consumption. For example, lighting science shows that in outdoor applications we can realize better effective brightness by applying the right CCT and spectral distribution. The result is better personal security relative to watts. Rea said that such a transition in approach by the lighting industry could lead to both increased profits for SSL manufacturers and a reduction in energy usage.
Improved light quality
Steve Lydecker, senior vice president of technology solutions at Acuity Brands, followed Rea in the Plenary session and reinforced many of the same messages on adding value in lighting (Fig. 4). "Lm/W, lm/$, and lm over life — those are the old boundaries of efficiency, economics, and endurance," said Lydecker. "But those table stakes have been met."
The lighting industry needs to move up the value creation curve to enrich and enhance our visual environment, according to Lydecker. But before he suggested how to make such a move, he used history to point out where the industry can't go.
"Our eyes are incredibly forgiving and tolerant to bad lighting," said Lydecker. He was referring to fluorescent lighting and the fact that it was only held to consistency within a 7-step MacAdam ellipse. He added, "As an industry we accepted fluorescent shortcomings" because it was cheap and easy to fix by changing a lamp.
LED lighting costs far more and is not easy to replace. Consistency must be engineered into the product. Lydecker said the industry needs to be held to a 1- or 2-step MacAdam ellipse.
Color metrics
One important measure of added value is color rendering. And again Lydecker pointed out shortcomings with current approaches, saying that color rendering index (CRI) was a simple measure and adding, "CRI is a one-dimensional measurement."
Adding the R9 red and other color samples has helped but not enough, said Lydecker. "CRI doesn't tell us how well we can discriminate between two colors in that matrix." Surprisingly, it was Lydecker that mentioned Gamut Area Index (GAI) rather than Rea since the LRC developed GAI as a second metric to be used alongside CRI as a measure of vividness.
Lydecker said, "GAI gives a much better feeling as to whether a scene will be vivid." CRI may be perfectly appropriate for residential applications, said Rea, but applications such as retail need a better way to characterize color rendering and lighting products that deliver on that metric. Lydecker said we need to "move lighting from the cost side of the equation to the value side."
Other angles to more value, according to Lydecker, include a move from static settings to dynamic experiences. And the examples may include moving light, but the point was really to use lighting elements and tunability to highlight parts of a scene. Again, saturated colors are key along with subtle CCTs in the white range.
Controls needed
Lydecker did say that light fixture design is not the only part of the equation, but that better tools (read controls) are required. Controls have consisted of dimmers and on-off switches in most cases. Slide dimmers provide 50 steps of mechanical resolution, and that's fine for dimming white light in low-end applications because the eye can barely detect a 10% level change, according to Lydecker. But for tunable color, Lydecker said, we need tools for color selection.
Lydecker showed an Acuity color-tunable product that was demonstrated at Lightfair last year and said the design includes multiple channels of control including a 16-bit color channel, an 8-bit channel that controls the distance from the black-body curve, and a 16-bit intensity channel. The industry needs to borrow from the display monitor space to see how such tunability is handled.
But we need more basic improvement in terms of dimming controls as well in professional applications; that may imply better LED lighting products. "We've all been pretty spoiled by a $10 dimmer we can buy at the DIY store and a filament inside a glass envelope," said Lydecker. That combination can deliver 2000:1 dynamic range. Of course, the eye has a dynamic range of 1,000,000:1, and we need to match controls to the eye and the application at hand.
The real action, according to Lydecker, is below 1% levels that the eye perceives as 10%. Above that level the applications are load shedding or energy management. But in spaces with daylight exposure or in conference rooms with projection displays, we need accurate dimming to 1%. In spaces such as theaters, we need accurate dimming to 0.1%, said Lydecker. And we need it without flicker — yet LED lighting is generally more prone to flicker at the lowest dim levels.
Now let's transition into some of the application-focused presentations. The topics are varied although the LED value proposition is a constant.
LEDs and solar
The combination of solar energy and LED lighting has come up at several recent Strategies Unlimited events. The presentation at SIL by Moneer Azzam, president and CEO of SolarOne Solutions, was the most comprehensive and provocative that we have seen. Saying that Azzam is boldly optimistic is an understatement. He said that solar plus SSL will cause a transition in outdoor lighting analogous to what the mobile phone has done to traditional phone companies. He asked, "Can this combination of technology loosen the grip that the utilities have on the access to the power side of the business equation?" At first his assertions may seem implausible, but Azzam is very convincing. And the reasons for using solar power are far broader than you might first expect; they could expand the applicability of LED-based lighting.
Back to the analogy, Azzam said that a mobile phone in 1973 cost $4000 for 20 minutes of talk time. That was pre-cellular, of course, and today voice is virtually free and wired phone companies are struggling to stay in business. Only in 2004 did solar lighting and LEDs combine in outdoor lighting. The $1/lm-h (dollar per lumen hour) effective cost and the 400-lm average output of a solar light were unusable. Today, the figures are in the range of $0.03/lm-h with an average output of 5500 lm in outdoor fixtures.
Ironically, it's not the fact that solar panels have magically broken through efficiency barriers or dropped in price that makes the technology a viable option. Instead, it's LED efficiency that makes solar viable in far more applications than in rural areas where there is no grid. We've covered a number of rural applications, but going forward we will find solar in areas near grid power.
Azzam specifically mentioned a roadway lighting project in Masdar City, Abu Dhabi (Fig. 5). The zero-carbon-footprint city actually already had massive solar farms but chose to use solar at the pole level for the roadway due to cost issues and to preserve the investment. Azzam said, "Moving the road was almost a certainty in the city's future." So installing a grid along the road could have been a waste of money.
Azzam also said installing wiring is incredibly expensive in areas such as Abu Dhabi, with sandy soil that ultimately requires multiple layers of protection for the power wiring. He said the installation cost can run as high as $100,000/km. The solar plus LED installation cost less than $85,000/km.
But it's not just exotic locales where you see a high cost in connecting to the grid. Azzam said that in cities such as Boston, Massachusetts, Washington, DC, and Toronto, Canada the cost to run wiring to a bus shelter is typically in the $6000 range. A solar-powered LED installation for a shelter can cost less than $1500.
You must also consider other sources of costs when contemplating new grid installation. Azzam discussed the rails to trails program in Massachusetts that is in the process of converting 450 miles of old railway beds into recreational trails. Installing a grid along those trails would have first required the decontamination of the soil. The cost could have been $100M to reclaim the soil.
Hard solar costs
But justifying the use of solar and LEDs together is not limited to instances where there is an infrastructure cost issue with the grid. Azzam said solar panels have dropped in cost, but prices are leveling. Still, he explained, "LEDs have already created a 20–30% reduction in solar cost" because the efficiency of LEDs means fewer panels are required.
Azzam showed an actual comparison of alternatives for a roadway lighting project in the United Arab Emirates. The twenty-year total cost of ownership (TCO) analysis considered the upfront cost of the luminaires, installation, and energy and maintenance costs. The solar-powered LED option was 10% lower in TCO than grid-powered metal-halide lighting. Grid-powered LED lighting came in a few percentage points below the solar alternative but the two were close. Furthermore, Azzam said that by 2016, the solar LED TCO will drop by 30% with additional technology developments. He said, "That's exciting and transformative."
Ironically, the cost picture improves when you consider the usage of adaptive controls. Azzam said a wireless network brings additional energy savings that offset the requirements for the battery, solar panels, and size of the pole required. Essentially the additional savings cover the cost of adding networks and controls, whereas the controls and network account for an added cost in grid-powered applications.
UV LED chase
Solar was also described as a potentially good match for developing ultraviolet (UV) LED technology in the presentation that followed. We'll explain why momentarily, but first let's cover the basics of UV LED technology. The UV spectrum is broken into UV-A (315–400 nm), UV-B (280–315 nm), and UV-C (100–280 nm). UV-A LEDs are relatively commonplace and used in curing applications, as we covered in a recent feature. The efficiency of UV-B and UV-C LEDs, however, drops rather drastically. SIL presenter Yitao Liao, a scientist at startup Rayvio Corp, said that under 300 nm, efficiency is in the 10% range.
But it's at the UV-C wavelengths that Liao sees the most potential for UV LED technology. The primary application would be disinfection that could be applied to water and perhaps surfaces or even entire rooms. Someday Liao foresees that a UV LED in your smartphone might let you disinfect your table in a restaurant.
Today, UV-B and -C application rely on mercury lamps. UV-B applications are mainly medical in nature where a bulky lamp may not be a huge issue. But the potential real-time-disinfection application for UV-C requires a portable source that isn't prone to breaking, has no hazardous material, and offers instant on and off capabilities.
As we covered in a UV article a couple of years back, Sensor Electronic Technology Inc. (SETi) is the only volume manufacturer of UV-C LEDs, and that company is using manufacturing techniques similar to those used in general-purpose LEDs with efficiencies remaining low. Other companies are pursuing exotic and more costly substrates. Liao said, "Growth in UV-C depends on a materials breakthrough."
Rayvio is a materials-focused company launched out of Boston University. The company is pursuing the use of molecular beam epitaxy to grow more-efficient UV-C LEDs on standard sapphire substrates. Liao didn't detail the company's progress. But he said a more-efficient UV-C LED could enable solar-powered water purification for the many regions around the world that lack access to clean water.
Camera flash advancements
Another exciting application for LEDs that isn't necessarily new — camera flash — has many new aspects to it. As we covered in our article on Strategies Unlimited market research, LED revenues tied to the flash market are growing. Yourii Martynov, director of strategic marketing at Philips Lumileds, examined the flash market in detail and explained that there may be even more opportunity for growth than most realize.
Phones and smartphones now serve as the primary digital camera for many consumers. Martynov said the megapixel race is over with no need to go beyond 13 Mpixels, and the image size doesn't drive the need for more or brighter LEDs. He did say the brighter LEDs would be welcome to extend the range of a camera sensor. The opportunity, however, is in the need for a second flash on the screen-side of the device for self-portraits and much better quality flash on the forward-facing camera.
Indeed, the light quality discussion prevalent at SIL, and the look of the LED, came to bear in the flash discussion. Phone vendors don't want LEDs with visible signs of phosphor because the flash can't impact the cosmetics of the phone. The vendors want better quality light that will allow their customers to capture better photos. That desire implies LEDs that can deliver consistent white color over angle at the very least, and perhaps even tunable-white flashes with excellent CRI and maybe even tunable color capabilities.
It turns out the color science that comes to play in the way humans see colors, which was discussed in the keynotes, also applies to the quality of images captured. Poor CRI in a flash can result in photos without vivid colors (Fig. 6). But Martynov said that really good photos need even more — a flash CCT that matches the ambient light in the area of the subject.
At a minimum, phone vendors want a neutral white that delivers the best compromise for a variety of lighting conditions. Professional photographers, however, place filters, including colored filters, on their flashes to match conditions. In the iPhone 5s Apple innovated a way to simulate that capability with a feature called TrueTone. The camera senses the ambient light and chooses from a 1000-scene palette to customize the flash profile. Martynov did not know, or did not reveal, how Apple implemented the feature but with- and without-TrueTone photos on Apple's website demonstrate the advantage.
Clearly, Martynov believes that there will be multiple LEDs in a phone flash going forward. It could be a mix of whites or perhaps a color array. The phone vendors have repeatedly shown that they will spend money on features that add value for the consumer.
Poultry production
Finishing where we started, we will now return to the subject of happy chicks. Brian Wilcox, president of Once Innovations, has his startup chasing unlikely customers in applications including both poultry and swine production, among other farming applications. Wilcox survived the Dragon's Lair at SIL and made a presentation in the market track.
The science that Wilcox presented was really amazing. Chickens have a far more complex visual system than humans with multiple receptors. The result is a broader range of sensitivity and the ability of tunable lighting to spur positive changes in production.
Red light, for example, can increase the growth rate of chickens and turkeys at the beginning of the rearing period. Moreover, it can reduce food consumption per egg laid while improving egg size. Green and blue light can lead to weight increase for broilers and higher prices at market. We hope to bring you more details on the science in a future article.
Of course, the advantage of LEDs with narrow spectrum comes along with energy reduction. LED lighting can deliver 97% energy savings relative to the high-pressure sodium lights used in many poultry operations. Wilcox said that one major complex realized a $650,055 benefit based on increased production and energy savings.
SIL 2014 began to reveal once again what can happen when the semiconductor industry applies its expertise in mass production of a tiny device such as an LED and, in doing so, enables a target application. Suddenly other industries find yet more uses for the low-cost enabling component. LEDs deployed in general lighting remain incredibly exciting and the primary driver of LED development. But leaving SIL the feeling that exciting applications are yet to come was unmistakable.
