Wednesday, June 02, 2010

Aptina History on Youtube

Aptina published a short video on its history on Youtube. Among other things, the video talks about Eric Fossum role, as it's seen from Aptina side:



Update: Eric Fossum wrote few historical remarks about JPL and Photobit times in comments. I'm copying them below:

This video was initiated by NASA's Hallmarks of Success project to document various successful tech transfer stories. The main tech transfer path was JPL->Photobit->Micron->Aptina so that is why Aptina features so prominently.

Roger Panicacci was a key chip design engineer in our group at JPL, a founding member of Photobit, and now a big honcho at Aptina and it was great to see him in this video (hey Roger!).

Besides Roger and me, the B&W JPL photo also shows Bob Nixon (retired), Barmak Mansoorian (now President of Forza Silicon), Bedabrata Pain (at JPL until recently - now in the Bollywood business), Orly Yadid-Pecht (now a Prof. in Calgary, I think), and others. Missing are team members Suni Mendis, Sabrina Kemeny, Junichi Nakamura (a Distinguished Visiting Scientist at JPL) and also Marty Agan who was a communications engineer at JPL and worked with us on a miniature wireless camera. I am sure I am forgetting other people known to readers of this blog and sorry about that.

I was not involved much at all in the making of this video and I can't say the video is 100% accurate, esp. as it is not aimed at image sensor technologists. It would have been good if it had mentioned other companies that licensed the technology directly from Caltech or indirectly through Photobit. And of course, APS was named by Tsutomu Nakamura of Olympus, and "3T APS" imagers (as they are now called) were around since the late 60's. What we claimed for inventions are spelled out in the claims of the actual issued patents but this would be way too much detail for this video.

At the time, in the early 1990's, CCDs were the indisputable king of imaging technology. The power dissipation of CCDs and associated electronics were enormous, and for space missions, CCD cameras were very bulky, power hungry, and prone to all kinds of failures. But their performance was/is extraordinary.

Our goal was to come up with a miniaturized scientific-quality image sensor technology that would maintain the performance but allow miniaturization. At that time, almost everyone (and I refer to the establishment of CCD guys) thought putting an ADC on chip was a BAD idea, much less integrating timing and control circuits, drivers, or digital processing. So, a CMOS based camera-on-a-chip (meaning camera electronics) was a radical idea. I did not know at that time of the notable work going on in Edinburgh or Sweden - but those efforts were definitely not geared towards image quality - they were geared towards low cost and minimal imaging quality (and in Linkoping, speed). They all used what I subsequently termed passive pixels to distinquish them from APS. (This was also what VVL and Omnivision used when they went into business. The Edinburgh and Linkoping teams definitely were part of making this whole camera-on-a-chip technology become ubiquitous today.

Anyway, after showing promising performance with intra-pixel charge transfer APS (first with MOS photogate and then JFET photogate (Pinned PD) with Kodak under tech transfer agreement) we launched Photobit because industry was moving at the speed of a snail. We also found that 3T APS worked pretty well for webcams and other applications and dropped the photogate approach for a few years until we could find someone who would make it in a timely way (Photobit continued to work with Kodak but that was sooo slow that it was dead before you knew it).

8 comments:

  1. eric fossum and aptina are the best in the biz. thanks for the post ISW. really appreciate this blog. it's the best source of cis info on the net (paid or free). :o)

    ReplyDelete
  2. why they are the best ? Please explain me ...

    thanks !

    ReplyDelete
  3. This video was initiated by NASA's Hallmarks of Success project to document various successful tech transfer stories. The main tech transfer path was JPL->Photobit->Micron->Aptina so that is why Aptina features so prominently.

    Roger Panicacci was a key chip design engineer in our group at JPL, a founding member of Photobit, and now a big honcho at Aptina and it was great to see him in this video (hey Roger!).

    Besides Roger and me, the B&W JPL photo also shows Bob Nixon (retired), Barmak Mansoorian (now President of Forza Silicon), Bedabrata Pain (at JPL until recently - now in the Bollywood business), Orly Yadid-Pecht (now a Prof. in Calgary, I think), and others. Missing are team members Suni Mendis, Sabrina Kemeny, Junichi Nakamura (a Distinguished Visiting Scientist at JPL) and also Marty Agan who was a communications engineer at JPL and worked with us on a miniature wireless camera. I am sure I am forgetting other people known to readers of this blog and sorry about that.

    I was not involved much at all in the making of this video and I can't say the video is 100% accurate, esp. as it is not aimed at image sensor technologists. It would have been good if it had mentioned other companies that licensed the technology directly from Caltech or indirectly through Photobit. And of course, APS was named by Tsutomu Nakamura of Olympus, and "3T APS" imagers (as they are now called) were around since the late 60's. What we claimed for inventions are spelled out in the claims of the actual issued patents but this would be way too much detail for this video.

    At the time, in the early 1990's, CCDs were the indisputable king of imaging technology. The power dissipation of CCDs and associated electronics were enormous, and for space missions, CCD cameras were very bulky, power hungry, and prone to all kinds of failures. But their performance was/is extraordinary.

    Our goal was to come up with a miniaturized scientific-quality image sensor technology that would maintain the performance but allow miniaturization. At that time, almost everyone (and I refer to the establishment of CCD guys) thought putting an ADC on chip was a BAD idea, much less integrating timing and control circuits, drivers, or digital processing. So, a CMOS based camera-on-a-chip (meaning camera electronics) was a radical idea. I did not know at that time of the notable work going on in Edinburgh or Sweden - but those efforts were definitely not geared towards image quality - they were geared towards low cost and minimal imaging quality (and in Linkoping, speed). They all used what I subsequently termed passive pixels to distinquish them from APS. (This was also what VVL and Omnivision used when they went into business. The Edinburgh and Linkoping teams definitely were part of making this whole camera-on-a-chip technology become ubiquitous today.

    Anyway, after showing promising performance with intra-pixel charge transfer APS (first with MOS photogate and then JFET photogate (Pinned PD) with Kodak under tech transfer agreement) we launched Photobit because industry was moving at the speed of a snail. We also found that 3T APS worked pretty well for webcams and other applications and dropped the photogate approach for a few years until we could find someone who would make it in a timely way (Photobit continued to work with Kodak but that was sooo slow that it was dead before you knew it).

    Well, guess I need to write this story down somewhere sometime. But, I am very happy to see this video and seeing how the technology has permeated so many areas of imaging. Like Hannibal said in the A-Team, "I love it when a plan comes together"

    p.s. Starting July 1st, I join the Thayer School of Engineering at Dartmouth College as a Research Professor which will be part-time at first incl. teaching. I expect to continue my other consulting work as well.

    ReplyDelete
  4. Thanks for the interesting story. I'm copying it to the post itself.

    ReplyDelete
  5. Eric, Are you going to continue to research in the same area?

    ReplyDelete
  6. JPL, Photobit, Micron, Aptina considered a success? They have been a success in some senses but they have not been a financial success even after 10 years of trying hard. A lot of money has been lost. A whole lot.

    ReplyDelete
  7. Anon, you are truly clueless about JPL. You also are clueless about Photobit's financials since they were never published. I can tell you for sure that Photobit Technology Corp was nicely profitable. When Micron bought Photobit, its value-proposition was quite close to Omnivision. In fact, we heard that Omnivision even proposed a merger with Photobit a few months before the sale to Micron, but of course we rejected that out of hand.

    I am guessing you are as clueless about Micron and Aptina as you are about JPL and Photobit.

    ReplyDelete
  8. Bottom line, Micron lost a lot of money. That is known.

    ReplyDelete

All comments are moderated to avoid spam.