Recently, I was invited to present at Image Sensors 2009. It was a great opportunity and not only because of the relative improvement in weather conditions traveling from Ottawa to San Diego. I presented, “The Competitive Technical Intelligence Perspective” which pretty much sums it up. I covered image sensor market trends, new products and closed it out by looking at R&D trends as measured through patenting activity. If you need to know more, a version is available elsewhere on this site. To maintain some advantage for people who paid to attend Image Sensors 2009, only a subset of slides are available, but I have included extra commentary so that it makes sense to anyone who wasn’t able to attend the live talk.

Enough about me. The focus of my conference observations is on TI - not that they were directly talking a lot about image sensors, but one of their product managers had some interesting things to say. Before jumping into that, I should make a quick comment about the attendance. The recession is definitely in full swing when it comes to travel budgets. Attendance was about one-third or less of previous editions of the conference. I just can’t get through a post without at least one depressing comment about the economy. Sorry.

TI has always been an analog company. Now they are taking that a step further - maybe becoming a metaphysical company. After all, they are no longer content with addressing the five human senses as machine interface. According to TI’s Ron Nag, now they want the sixth sense. To help them get there, they are bringing up a 300mm analog fab in Richardson, Texas.

TI sees opportunities in your kitchen and your bathroom. No, they are not considering branching into home renovation to take advantage of our Canadian government tax incentives meant to lift us out of the recession. Today, we are all dependent on Google for our information, so it’s logical that we need a convenient way to access the internet when we are cooking. As Ron Nag put it, though, countertop space is at a premium. Although no specifics were provided, TI is looking at ways to interact with the computer without traditional input like keyboard or mouse or even touch screen. Perhaps it will be Vulcan mind control.

Well this was an image sensor conference, so I’m sure you can guess that they are really looking at using cameras to register hand gestures as input in much the same way demonstrated by Microsoft’s Project Natal.

As for the bathroom, do I dare speculate? I suppose building in a screen for Internet or an e-reader might help reduce some of that unsightly magazine and newspaper clutter - dead giveaways about where you spend your time reading. I’m not suggesting that putting a screen in the bathroom will somehow provide a cover for your activities, but employing the sixth sense means no reading materials get flagged for a trip through the bathroom. Then again, if your household is anything like mine when it comes to replacing the roll, extra paper can come in handy. Please keep the Sears catalog off-line.

Whatever the case, TI sees big opportunities for sensors and their analog products coming out of Richardson.

TI is becoming more of an IP company. Not just the licensing type with lots of lawyers and reverse engineering companies. Now they are selling their know-how in DSP algorithms to customers as part of a complete solution. TI also strives to give the large community of ARM core developers full access to the peripherals and other IP blocks on the DSP to create a development environment more akin to an API that pure software developers are used to.

It was not the TI rep, but maybe it was the influence of their sixth sense. In any case, one of the other conference presenters referred to me as, “Dan.” You may recall from that long defunct SemiSerious blog that the same thing happened to me after another presentation at Image Sensors given in London last year. I guess it really is time to finally change my name (or stop attending this conference).

There were a few other bits of interesting if not downright potentially useful information picked up in San Diego:

Consolidation is not happening in Image Sensors. (Tom Hauskens, Strategies Unlimited)

Consolidation IS happening in image sensors within specific market segments like cameraphone. (Tom Hauskens, Strategies Unlimited)

Shipments are growing but revenues are not. Declining margins will push many players to the margins - into niche applications or out of business. (Shri Sundaram, Toshiba America Electronic Components)

Expect five to 12 players with only three holding a large chunk of the market. (Shri Sundaram, Toshiba America Electronic Components)

Data analysis and decision needs to move to the edge of the network at the sensors (for security and similar applications) to avoid large data streams over communication channels. (Ken Johnson, Solidica)

Yield at the module level is worsening as pixel size is reduced. Mechanical tolerances are not able to keep up with pixel scaling. Microlens alignment is also more difficult. DxO is now applying their extended depth-of-focus technology to the task of improving camera yield. The French company now applies its algorithms to apply corrections without performing costly calibration steps. (Frederic Guichard, DxO Labs)

There you have it. I probably could have shared more, but I feel like this post is already a week late. (My presentation was on October 14.) Image sensor technology is a fascinating field. I only wish there was more time available to spend on it