Interview with Ed Motamedi
Originally published in the MEMS Investor Journal, and reprinted here by permission.
While MOEMS, the combination of micro-optics and MEMS, have been slowed down by the telecom downturn in the early part of this decade, these technologies are now making a comeback. In addition to telecom, MOEMS applications include sensors, projection and mobile systems and devices.
We recently spoke with Dr. M. Edward Motamedi who is the author and editor of a recently published MOEMS book and the founder of Revoltech Microsystems. Dr. Motamedi is one of the top experts in the field of MOEMS and is a Fellow of SPIE.
When MOEMS first introduced and what were the original applications?
Motamedi: During 1991-1993, we started to use the acronym of MOEMS for "Micro-Opto-Electro-Mechanical Systems". This was to distinguish between optical MEMS which also include bulk optics. In 1993, I introduced MOEMS as the powerful combination of MEMS and micro-optics in an invited talk at the SPIE Critical Reviews of Optical Science and Technology conference in San Diego
The original MOEMS applications reported in that presentation were laser scanners and dynamic micromirrors for AO applications. However, at that time, and possibly a decade before, in the early 1980s, micromirror devices were already under development at TI where they eventually became the foundation of DMD. During the 1990s, the first commercial realization of the DMD became feasible. DMD was a truly MOEMS device.
How was the market for MOEMS affected by the troubles in the telecom industry in the early part of this decade?
Motamedi: When the FCC reported that the telecom industry earned $270 billion in revenue during 1999 including more than a 30% jump in wireless infrastructure and sometimes more than 60% in other categories, that was the final report of the telecom boom in the 1990s. Back then, there was an inflated view of telecom which generated a globally false vision of this industry and resulted in the telecom bubble bursting in 2001.
In 2000 and before the telecom bubble burst, since bandwidth and faster throughput were major requirements for the high growth rate of the telecom industry, and because "all optical networking" was promising to fulfill these requirements, many venture capital firms rushed to invest in many companies which actually did not have proper knowledge of the field.
Since the increased demand for bandwidth and throughput can really only be fulfilled by MOEMS, the market for these technologies was unfairly affected by the telecom bubble burst and many high level innovative ideas for development of bypass switches, variable optical attenuators (VOA), optical tunable filters (OTF), and high density wavelength division multiplexing (HDWDM) were lost during the overall downturn in the telecom industry.
Even though many of these ideas are coming back to life these days, some have been set back that so far that many experts in the field have had to make a career change.
Which applications have survived and are on the market today?
Motamedi: One of the original applications of MOEMS, laser scanners, has survived. It is now commercially available from Intermec Technologies Corporation. DMD has been further developed at TI and was introduced to market as DMD arrays. In 1996, DMD development progressed rapidly to a sophisticated low-cost chip and became the foundation of DLP technology which is currently used by several manufacturers worldwide and covers approximately 50 percent of global market share of all front projection TVs.
What are the current challenges?
Motamedi: There are some fundamental challenges for process development of MOEMS chips. Optical surfaces of micromirrors have to be flat to a fraction of light wavelength to be able to use the chip for high resolution imaging. Depending on which light wavelength is used, mirror surfaces should have total reflection and this requires coating the mirror to achieve high reflectivity. Another challenge is developing a CMOS compatible process to be able to integrate the electronic circuit with the MOEMS chip. Another process challenge is increasing the wafer size - this is specially challenging when bulk micromachining is preferred to surface mismatching. Yet another major challenge for MOEMS devices, is packaging. Today and even through the current decade, MOEMS chips with optical window will demand innovative ideas for packaging and challenges to make sure that the design is cost-effective.
Which startup companies are now especially promising? What are their products?
Motamedi: Startup MOEMS companies were driven initially to support the telecom industry using some innovative ideas to revolutionize the needs of high bandwidth and throughput in all-optical networking by filling the gap in development of WDM, DWDM and HDWDM to reach the manufacturing stage. Even though some of the startups had attracted investment with non-telecom products such as laser scanners, micromirror arrays, and microdisplay, the telecom boom during the late 1990s, forced them to update their business plans to stay alive. Some companies like Microvision, MEMS Optical, and JDSU have survived. Microvision products are based on microscanners, MEMS Optical products are micro-optics components, and JDSU products are mostly based on telecom.
Which large companies have projects in this area? What are their projects?
Motamedi: Texas Instruments is the leading company in manufacturing consumer electronics like digital TVs and digital projectors with the applications of DMD and DLP. These applications may be called optical MEMS but in fact, since DMD and DLP chips are made by the marriage of MEMS with micro-optics, the end products are actually MOEMS. TI is the largest company producing MOEMS with revenues reported close to $1 billion. The other large companies manufacturing MOEMS-based products are Hewlett Packard with revenues of $500 million, Robert Bosch with revenues of $374 million.
Which are the leading groups for MOEMS in academia?
Motamedi: Leading MOEMS groups in academia are UC Berkeley, Boston University, Stanford, Fraunhofer Institute and MIT. In addition, the following universities and institutes also have substantial efforts in MOEMS; these are listed below in alphabetical order:
University of Albany, University of Delaware, Drexel University, University of Jena, University of Michigan, Montana State University, University of North Carolina, UC Santa Barbara, Sariyer University, University of Tokyo, and University of Wisconsin.
Which MOEMS applications are likely to be commercialized over the next 3-5 years?
Motamedi: In my opinion, telecom will be back to life during next decade since it is one of the drivers for our economy. Companies will continue their efforts to commercialize MOEMS components for the telecom industry over the next 3-5 years. Some of the specific components are:
VOA devices: with challenges to 0.1 dB step attenuation, crosstalk better than 75 dBm, and switching speed in ms range.
Optical "NxN" matrix switches: achievable through optical beam interception and deflection using beam splitters with fast actuators. Increasing the value of "N" and improving the reflectivity of the beam splitters close to 100% are challenging.
Tunable Filters: Components like DWDM and HD-WDM are extremely demandable for being commercialized.
In summary, could you comment on the lessons learned and future trends for MOEMS?
Motamedi: Yes, we learned lessons from the telecom "boom and burst". Here are what went wrong - investors and government funds were poured onto overnight startups without really evaluating the companies' business plans and capabilities. Large companies started competing to acquire startups with prices that were orders of magnitude higher than fair values. As the result, respectful MOEMS companies lost support and many ground-breaking programs were lost.
The trend for MOEMS during the next decade will be in revolutionizing photonics systems with breakthroughs in telecom, microdisplay and consumer electronics fields. Emerging new services like HDTV and web-based video conferencing will demand increasing bandwidth and transmission rates on the order of 10Mb/s. These requirements are only achievable by MOEMS, the combination of two microtechnologies of MEMS and micro-optics.
Dr. M. Edward Motamedi has a Masters and a Ph.D. from Northwestern University in Electrical Engineering and an MBA from Pepperdine University with a focus in Executive Leadership. He is currently an executive and founder of Revoltech Microsystems where he has guided technology development of MEMS and micro-optics and now is leading the transition to manufacturing.
Dr. Motamedi is the founder and past chair of SPIE MOEMS-MEMS Symposium and now serving as the chair of the symposium steering committee. He has also chaired several SPIE conferences on MEMS and MOEMS, was the editor of eleven SPIE proceedings, and has developed three short courses in MOEMS. Dr. Motamedi has moderated several technology panels and workshops including the latest one on January 25, 2005, a panel entitled "Technology, Market Review, and Roadmap extended to 1010". For more than a decade he has been promoting the MOEMS field as a microtechnology of merging MEMS with micro-optics.
Dr. Motamedi is the editor and author of the textbook Micro-Opto-Electro-Mechanical Systems which was published on April 2005. He has more than 100 publications including numerous invited papers and plenary presentations. Dr. Motamedi is a senior Editor of the Journal Micro/Nanolithography, MEMS, and MOEMS. He is a senior member of IEEE and serving as Vice Chair of IEEE Buenaventura Section and as a senior member of IEEE Los Angeles Council. He has 16 national and international patents and is a recipient of SPIE 2005 Lifetime Achievement Award. He is a Fellow of SPIE.