Kodak exits opto and ends an era

It seems like the end of an era: Kodak is selling its CCD operations and its image sensor patents. It had been making CCDs since 1975, one of the early companies to make them, but waited until 1989 to sell them externally. Kodak had a number of firsts, including the first megapixel sensor, in 1986.

Then CMOS image sensors took off.CMOS sensors were conceived early on, but the lithography was too poor at the time. Omnivision and others brought it to life in the 1990s. Kodak tried several times to break into that product line, but it never worked out. Kodak teamed with Motorola in 1997 on CMOS image sensors. In 2004 it acquired National Semiconductor’s CMOS image sensor operation, for about $10 million in cash. Kodak even had deals with IBM and TSMC to manufacture the sensors, and some clever technology. But it wasn't enough.  

In our 1997 market report, we estimated that Kodak was the leading producer of image sensors outside of Japan, with $38 million in sales and under 6% market share. By the time of our 2009 market report, the image sensor market had grown 10X, but Kodak’s sales were stuck for years at about $80 million. Then in April it sold hundreds of patents and patent applications to Omnivision, for $65 million. And now it’s selling the CCD facility and its 200 employees to  Platinum Equity, a private equity firm.

In a way, kicking out the CCD business has little in common with the rest of Kodak’s problems. The operation being sold still makes high performance CCDs for high-end professional and scientific applications--some of it is really amazing stuff. And over the years a lot of companies have handed off their image sensor operations. For example, Pixel Devices International was sold to Agilent, which became Avago, who sold the image sensor operation to Micron, which spun it off as Aptina. And of course, Kodak is still huge into imaging, and that's photonics too.

It’s just the business getting older, but Kodak had been a classic example of a U.S. company deep into optoelectronics--that is, the actual making of the chips. No more.

LED drivers--a $2 billion photonics market

With all of us opto folks going gaga over the $10+ billion LED market, stop and consider that the LED driver IC market is a sweet $2 billion, and growing at 12% compounded annually. And I'm going to say it: drivers is a photonics market too.

OK, I said it. Electronics is photonics too. I'm stretching things a bit, since the suppliers of LED driver ICs are companies like Texas Instruments, Maxim, Analog Devices, and Macroblock who don't know or care about photons. They do know a lot about hand-crafted analog circuit designs and specialty fab processes that enable circuits tolerant to high-voltages--the kind that drive long LED strings in display backlights.

But good LED design optimizes the entire circuit for efficiency, reliability, LED uniformity, and many other specs. We call the circuit--minus the LEDs themselves--the driver. It may include zero, one, or multiple ICs for the purpose.

Opto people, like myself, tend to think that there is nothing interesting in the system apart from the quantum mechanics of electron-hole recombination and fancy MOCVD epitaxial growth.  But when product designers take the electronics for granted, system performance is notoriously terrible, and that's bad for the whole LED industry. Likewise, electronics designers tend to take the LED for granted, but LEDs are requiring surprisingly novel and sophisticated circuits.  The only way to achieve widespread LED lighting is if electronics designers innovate enough to meet cost and performance goals.  Fortunately, there are those out there who can. Look for example at companies like Exclara, iWatt, Luxera, and Lynk Labs, to name a few.

The boundary between electronics and photonics is also fuzzy for lightwave transceivers. The laser and detector in a transceiver are typically very cheap, so much of the value is in the electronics: driver and receiver, clock recovery, and so on inside the module, not to mention all the higher level routing and control elsewhere on the board.

In imaging, it is even more dramatic. The detector array is sophisticated, but the image processing electronics takes it further, correcting optical limitations and even adjusting focus after the fact. The point is not that the electronics helps the optics, but that optical science actually resides in the electronics, often on the same chip as the sensor array.

I'll get back to the LED driver market again, but for now, remember: Electronics can be photonics too.

A big optics/vision opportunity: service robots

Looking for a new opportunity in optics and vision systems?   Check out this new  market report on vision for service robots from my colleagues at Vision Systems Design.  This is a market set to take off.  To give you an idea, with industrial robot unit sales in the tens of thousands per year, service robots could potentially sell in the millions.

Most robots today are not the futuristic kind we remember from the Jetsons or the somewhat creepy Actroid kind commercialized in Japan..  An industrial robots today is basically just factory automation with an articulating arm that makes it seem like a robot. 

A service robot is more like the more futuristic version--mobile, uncontained, and diverse--but not trying to act human, like some insecure, fawning android.  More precisely, it operates semi- or fully-autonomously to perform service functions, excluding manufacturing.  An industrial robot can be a service robot too, if it meets this definition. 

Examples of service robots include: UAVs, explosive or hazard disposal, automating cow milking, driver assistance, inspection and maintenance of hard-to-reach places, medical rehabilitation, surgery, and scientific exploration.  The UAV is the biggest market opportunity, becuase of the sophistication involved.  There are many smaller, fast growing segments.

This is a big deal for photonics because most service robots requrie machine vision of some kind.  This means the use of structured light (like what is used in the Microsoft Kinect), time-of-flight (like what is used in virtual keyboards), LIDAR, and so forth.  This has to be fused with other technologies, like GPS, radar, sonar, and inertial guidance.  For more sophisticated robots, simultaneous localization and mapping (SLAM) is critical to build maps of unknown environments or to update maps within known environments, while at the same time keeping track of the current location of the robot.

The technology is still emerging and remains to be worked out.  That means lots of hardware and software, and pretty deep stuff.  Imagine that the system doesn't necessarily need to "see" things the way we do--it just has to get the information it needs from its sensors.

For more information on the report, click here.

The Decade's 5 Best Market Trends in Photonics

Well, it'over: the decade-that-could-not-be-named (2001-2010). Time for a list of the Decade's 5 Best Market Trends in Photonics:

The triumph of flat displays. Remember the see-through iMacs? Don't CRTs look soooo 20th century now? And how about the touch screens for smart phones? They don't just make a nicer phone. They transform how we live. This is #1 because of the sheer size of the display industry, and its impact on everything else.

Cellphone cameras relaunch the image sensor market. The first cameraphone showed up in 2000, in Japan. There are now over 1 billion mobile phone handsets shipped every year, and most have cameras. That volume drives lots of other applications. And the quality! Again, transformative, and billions in new photonics revenue.

The Green Revolution: LEDs and solar. High-brightness LEDs are all about taking an old technology and improving the brightness to do some new things--a marketing VP's dream. This decade saw LEDs in mobile phones and TV backlights, but the talk now is about LED lighting taking over the world. And how about that solar market! Many investors have lost money in solar companies, but they keep coming. Somehow, we all want to be part of it. (I was too, back that was back in the 1970s.)

Molecular imaging and all other thing biophotonic. The average person on the street doesn't know it, but photonics is making a huge impact in biomedicine, from diagnostics to therapy. Optical molecular imaging is my favorite because of the promise it brings in finally solving some difficult and costly medical challenges. But there is also mid-IR spectroscopy, ultrafast surgery, OCT, and many more.

IPG and the fiber laser. Several companies had fiber laser products in the 1990s, but IPG Photonics gets credit for making it a big name in materials processing today, and the 5th largest maker of non-diode lasers. Ha! No one is laughing now. The fiber laser is one of the laser types you would design if you could pick only one, and if you had the materials you have today. The diode laser is the other.

There you have it. Next time if I get to it, the Decade's 5 Un-Trends in Photonics.

Microsoft Kinect is a win for lasers and imaging

Bravo to Microsoft this week for launching a new market segment for lasers! The company finally launched the Kinect, a laser-based accessory that allows Xbox360 players to use their body as a remote control--gesture recognition. The lasers illuminate a field and an imaging system interprets it for the game console.

The reviews so far indicate that Microsoft is on to something here. It is a step up from the Nintendo Wii remote controller and way beyond the Sony PlayStation Eye, neither of which used laser projection to create the image. The main complaints so far seem to be that the Kinect is understandably a little clunky yet, and it doesn't work well in crowded spaces, like dorm rooms.

The system needs single-mode lasers to get a strong enough return signal to do the processing. But, they have to operate in the near-infrared to be invisible and at low enough average power to be eye safe. Kinect uses a structured light technique, which is to say that it interprets the distortion in an image created by the object. When the object moves, the field is distored accordingly.

The project was born as Project Natal, from two Israeli companies, Prime Sense and 3DV Systems. Microsoft bought 3DV in 2009, and this week it bought Canesta. Canesta is known for its laser-based 3D imaging system based on the time-of-flight technique. It works much like radar to recreate a 3D form. Years ago Canesta used the technique to read finger strokes on a virtual keyboard. It worked, and although I prefer an actual keyboard for touch typing, it may be useful as an alternative to touchscreens. The product is sold through a joint venture called Celluon, created in 2004.

The Kinect technique is less expensive than the Canesta approach. Maybe Microsoft has other plans for Canesta. Or it wants Canesta's expertise in imaging. Or maybe it wants to keep it away from competitors.

Any way you look at it this is all good news for lasers and imaging engineers. It's the beginning of a new market segment.

New report: OCT has a good 2009

How good is the OCT equipment market these days? It grew in 2009, despite the recession, and that's in units, not just exchange rate adjustments. That’s how good it is. What’s more, we’re expecting 20% compound growth through 2014. This is from our latest OCT market report, now officially released.

If you don’t know what it is, OCT is a great imaging technology. OCT systems use advanced optics to construct micron-scale cross-sectional and 3D images in real time. The technique is non-invasive and can be used in vivo, such as to examine the inside of the human eye. The business is now entering a new phase, beyond ophthalmology to new specialties, such as looking inside arteries for cardiology. Each of these applications has the potential to be as big or bigger than the sales today in ophthalmology.

OCT may sound arcane, especially if you have to pronounce its full name. But this is big stuff, with equipment sales now in the hundreds of millions of dollars per year. Much of the activity has been driven by a shift in the technology from time-domain systems to faster Fourier-domain systems, which are also not as limited by patent protection.

Carl Zeiss Meditec no longer holds the majority of market share, even as its own revenue has grown. At least 21 other companies are developing and/or marketing OCT systems in 8 medical specialties, as well as in R&D and industrial applications. Many more companies supply the sources, detectors, and related components that enable OCT systems and applications.

Oh, and did I mention? Our market report is the only comprehensive report on the topic, a follow-on to an earlier report, also sponsored by PennWell. Kudos to our good friend, Greg Smolka.

Consolidation--Part 3: Image sensors as a case study

I can't think of a photonics market where someone doesn't frequently say, consolidation is a-comin'. The image sensor market is no exception. So when we finished our new image sensor report, I wanted to see how it looked over the 13 or so years that we've tracked that market. And you know what? The more things change, the more they stay the same.

At the high level, there hasn't been one iota of consolidation in all those years, despite the fact that the image sensor business has grown about 10X in that time. The first chart shows a simple way of measuring the consolidation. It shows that the top 5 suppliers have had about the same overall share over many years.

I should note here that the membership in the Top 5 has changed during that time. That is, the market share is not static. It's just that there is no sign of a move toward consolidation at that level.

Individual segments, on the other hand, are often highly consolidated. For example, the sales of image sensors for security cameras, for optical mice, and for certain scientific and professional applications are held by just a few players in each case. In fact, this is typical of photonic products, where the overall category is really a hodge podge of subsegments, and few, if any, suppliers can ever achieve significant market share overall.

The next chart shows how the number of suppliers has changed over the years. Far from consolidating, the number of suppliers has increased since the commercial success of CMOS image sensors in this decade. In fact, just as one supplier is acquired or exits the market, another pops up.

So, when referring to consolidation, be sure to mention whether you are referring to a narrow segment, or the overall market. Consolidation in narrow segments is common, while consolidation in the overall product group is unusual.

Image sensor market enters a more cyclic phase

It took the recession of 2009 to do it, but the image sensor market will finally take a dip after many years of steep growth. And, not only has the market contracted for the first time since we first started tracking image sensors in 1996, but a rockier period waits ahead.

Image sensors have a long history, dating back to fax machines and video cameras, but the recent steep run is largely thanks to the cameraphone. The cameraphone provided a convergence of factors that drove revenues to steep double-digit levels: a simultaneous swelling of handset sales worldwide, growing adoption of cameras in handsets, and rapid migration toward greater pixel counts in those cameras.

Now handset sales are taking a breather, dropping an expected 10% in unit sales this year. Other segments are also seeing declines, of course, and unit prices for the image sensors are relentlessly competitive.

The figure shows the long steep ride on a log scale, from our new report. The log scale makes it easy to see the change in growth rate (a chart with a linear scale is in our press release). What's remarkable, perhaps, is that the market decline this year won't be more severe than it is already.

We still expect single- and low double-digit growth in the recovery, and there are still many new opportunities for growth in unit sales beyond that. The two most notable are automotive and webcams. But, it gets more and more difficult to drive revenues up when it’s already near the $7 billion mark. Without the fortuitous convergence of rising cameraphone sales to drive the market forward, cycles will become more prominent.

By the way, it's not that the suppliers have had an easy time of it up to now. (Sound familiar, anyone?) But that will be a topic in a later blog.

InterSolar, PW, and gadgets are in; big Semicon West tools are out

SEMI's still newish InterSolar show outshined it's collocated Semicon West show this week, and that may be a sign of the future. InterSolar now has 3 floors of Moscone West, has all layers of the supply chain represented, and was simply more exuberant. The building has big windows. It feels more open. The solar industry this year is down, too, and there are still fewer vendors than at Semicon, but it just seemed brighter at the InterSolar show.

In contrast, Semicon is underground at Moscone North and South. It was cheerful enough, considering the downturn and the lack of windows, but the party was at the solar show. And judging from the signup map for next year, the North and South Halls will be even emptier. In fact, Photonics West is now clearly bigger than Semicon West--at least in exhibitors and floor space, and maybe attendees too. (Rumors that Semicon will be squeezed into the South Hall turned out to be false. They are putting all the wafer processing booths in the South Hall. The North Hall will have everything else.)

To be fair, no one was expecting Semicon West to be much of a party this year. After all, SEMI just announced that tool sales will drop 50% this year to the unspeakable low of $14 billion. (It was $43 billion in 2007.) But it's notable who is showing there nowadays, or rather who isn't. Semicon is a tool show, but Applied Materials, KLA Tencor, LAM Research, ASML, and many other major toolmakers don't have conventional booths anymore. For example, Applied and KLA only had meeting rooms at Semicon, while they showed their solar tools across the street at InterSolar.

Semicon West is now really about the gadgets that the major vendors attach to their systems, the materials they use, and R&D lab equipment. This means that there are suppliers for everything from microscopes and instruments to encoders and bearings. It's a good show for this kind of product development and lab stuff, and some of the specialty tools that are used in North America.

But this year, even that was down. In lasers, Cymer, Gigaphoton, Coherent, and JDS Uniphase--each one catering to the semiconductor industry--all didn't show. The laser companies that I saw were Deep Photonics, Eolite, DPSS Lasers, Innolas, Jenoptik, Quantronix, and Rofin-Baasel. IPG and Newport showed their lasers at InterSolar.

SEMICON is still the biggest fab tool show of the year for North America, and a very important one. One instrument vendor told me that even in a year like this they expected to get some sales from leads at SEMICON. New tool development still goes on. But it's a shadow of its glory days, about 15 to 20 years ago. The SEMICON shows in Asia are much weightier on the big tool side. And that makes sense, since as much as 75% of the world market for semi tools is in Asia, according to SEMI.

And, the industry has grown so big that there are other, more specialized shows to choose from. Like, for examplek, the lithography people like the SPIE Advanced Lithography meeting held in San Jose. There are meetings like this for every sort of nuance you can think of.

Photonics West is also a gadget show aimed at product developers and lab workers, like Semicon, but it spans more industries, particularly healthier ones, like biomedical and security. And it has a cross-cutting technical conference too. (And don't forget the Laser Focus World Marketplace Seminar, collocated with Photonics West every year!)

I think it's a sign of the times. Solar is in. Photonic gadgets are in. Semi tool gadgets are still in. But the days of the big tool show in North America may be over.

Military photonics up, machine vision down

The military and security market is one of the few that's faring reasonably well these days. Compare that to machine vision, which is being hit hard with all the other capital equipment markets right now.

Here are two companies that are pretty good proxies for the sales in those two markets: FLIR and Cognex. Both are heavily focused on imaging, but for different markets. The figure shows their revenues since 2006. I didn't adjust for some some minor acquisitions that each of them made, but it doesn't seem to matter much.

FLIR's sales are hardly off it's steady upward climb. Its internal sales are much more complicated. Its government sales and commercial vision sales were up, year over year, for military and security applications. Thermography was down due to falling demand and exchange rates.

Compare that to Cognex, which specializes in machine vision and automation. One would hope that machine vision is one of those markets that may still defy the recession somewhat, particularly from customers looking to improve costs during hard times. But alas, machine vision equipment is capital equipment too, and customers are trying hard to keep capital spending low for now.

If the rest of the year were flat for each company, FLIR's year would end up 1% over 2008 and Cognex's year will be 30% under 2008. That's right in line with our view of their respective markets. The military sector is still going strong, but is at or near a long-term peak. We expect the manufacturing sector to be down about 30% in 2009, but sales will be better in 2010.