GETI/GLOCOM Platform Joint Seminar

Progress and Changes in MEMS/MST Sector in 2002

Summary by GETI Staff

Advancements in MEMS (micro-electro mechanical system) and MST (microsystems technology) since the 1990s have led to the proliferation of related conferences and society meetings and workshops. Over the past several years, the increasing interest in these fields has led to the development of key areas of specialization. For example, we have seen many conferences and workshops on the development of MEMS/MST for optical networking and display applications (Optical MEMS), for the handling of liquid samples on a chip for biochemical analysis (micro total analysis system, or "μTAS"), for power generation applications (Power MEMS), and even for the commercialization of products based on these technologies. Ms. Waga has attended and spoken at many of such meetings throughout 2002. Ms. Waga reviewed the year in MEMS/MST activity for 2002 and discussed some of the major issues in this emerging area.

According to Ms. Waga, 2002 was a year of big change for companies engaged in MEMS/MST development and commercialization. The key change resulted from the sorry state of the telecommunications sector which continued to slump in 2002. As a result, a number of optical MEMS start-ups went out of business, and even bigger companies gave up optical MEMS-based product development projects and laid-off engineering personnel. The impact on MEMS foundries and design tool suppliers was also severe, leading to more personnel cuts. In the summer of 2002, the acquisition of Cronos Integrated Microsystems, a unit of the Canadian optical component company JDS Uniphase, by MEMSCAP of France in a stock deal of $10m (compared with the $750m merger deal JDS made with Cronos in April 2000) came as a surprise. The reduction in market value of Cronos became a leading market indicator of the weakness of the optical MEMS area. The sharp decline in company value (from $750m to just $10m) illustrates the extent of the over-valuation problem. Telecoms component manufacturers and MEMS foundries suffered the consequences.

The news about Standard MEMS's bankruptcy was another blow to MEMS in general, because the company was regarded as one of the most promising MEMS companies earlier in the year. While these changes where derived mainly from the drastic reductions in capex by the telecommunications sector, these events also seemed influenced by the rapid expansion of MEMS foundry capabilities in Taiwan. Walsin Lihwa invested $70m in its MEMS foundry capacity development. Asia Pacific Microsystems, which was founded in 2001 and raised $50m in the first round of financing, bought production facilities from Winbond, a large IC foundry in Taiwan, and started foundry services using 6" wafers. These firms have won orders from customers abroad and are very aggressive in business development. The expansion of MEMS production capacity in Taiwan has put pressure on firms with higher costs and led to lost orders to the Taiwanese competition. These deflationary pressures may be good for customers who could acquire cheaper MEMS devices, but due to the lack of growth in demand companies are not benefiting to the extent that they would have absent the aforementioned IT sector troubles.

In regard to international cooperation, Ms. Waga mentioned that the successful combination of design expertise from the U.S. and manufacturing expertise based in Asia led to the successful growth of MEMSIC, a company that produces MEMS-based accelerometers. MEMSIC was established in 1999 and is located in Andover, Massachusetts. MEMSIC has retained Taiwan Semiconductor Manufacturing Co. (TSMC) in Taiwan to produce and run the wafer foundry for the CMOS process, with Nantong-Fujitsu Microelectronics in China acting as the product packaging foundry. MEMSIC has a wholly owned subsidiary, MEMSIC Semiconductor (Wuxi), LTD., in Wuxi, China. (Wuxi is 80 miles away from Shanghai and it is at the heart of a fast growing Chinese microelectronics sector). According to MEMSIC, "By establishing its own manufacturing in China (a wholly owned subsidiary) and maintaining development in the United States the Company has the best of two worlds. The Company is more cost competitive due to lower manufacturing costs and has a higher return on capital investment than all of the major competitors and still has access to new technology developed in the United States". If this business model proves successful and is followed by other firms, competition in the high-volume MEMS device market will be very tough against those who produce in countries where manufacturing costs are significantly higher. The prospect of the aforementioned occurring was a point of great concern by those Japanese managers in attendance at the seminar.

The positive news for the MEMS/MST sector in 2002 was the continued expansion in R&D funding, according to Ms. Waga. At present, over $2b/year is being spent on nanotechnology initiatives all over the world. The increased funding has provided incentives for researchers to conduct work at the nanometer range or to work on radically innovative methodologies that will lead to nano-size particles assembling themselves (i.e. "self-assembly") into functional elements. This trend from MEMS to NEMS will likely continue mostly in academia and in industry, to a lesser extent, in the years to come. In regard to the above, Ms. Waga informed the attendees that DARPA (the U.S. Defense Advanced Research Projects Agency) had reportedly earmarked $124m for MEMS, microsystems and nanotechnology R&D in 2003. DARPA has been instrumental in supporting the development of MEMS technology and industry in the U.S. In Europe, the EU's 6^th Framework Program for Research and Technological Development will be initiated in 2003, providing continued support for MST development. It is estimated that the EU supported over 160 projects related to MST in the last 3 years. Each project involves several research groups in academia and industry from various countries. Such continued public investment will assist scientists and engineers in their efforts to demonstrate the technical feasibility of innovative ideas and help companies to bring new products to the marketplace.

Speaking of technology trends by application, Ms. Waga discussed the strong R&D support of micro power generation technologies globally in 2002. DARPA has played a pivotal role in promoting this emerging field of technology, and the U.S. is the world leader in research into micro gas turbines, micro rotary engines, micro fuel cells, and micro propulsion. German research institutes and companies are also very advanced in micro fuel cell R&D. For instance, the Fraunhofer Institute has demonstrated micro fuel cell prototypes for use in notebook computers and camcorders. In Japan and Korea, several companies have announced various types of micro fuel cell prototypes in 2002 and are aiming to commercialize such systems in 2004-05. Similar work will be publicly promoted and supported in Taiwan. Other application areas that look promising for MEMS/MST developers include areas within the life sciences and wireless communications. The worldwide shipment of MST products for medical and biochemical applications is expected to grow from $2.4b in 2000 to $7.4b in 2004, achieving a CAGR (compound annual growth rate) of 32.5%, according to the independent consultant Roger Grace Associates. Major challenges in the biomedical MEMS/MST sector, however, include the long approval process by authorities and safety concerns. In the case of devices that directly interface with the human body, such as implantable drug delivery systems, the technology developers must find ways to sustain prolonged R&D and clinical trial work. This will mean that the development of these technologies will require "patient capital."

In the wireless communications area, Intel and IBM announced research efforts related to MEMS in 2002. Some companies started shipping RF MEMS switch samples for civilian communications and instrumentation applications. Advocates of RF MEMS say that it is the next killer application for MEMS/MST because the new handset market has strong growth rates. About 400 million handset units are sold worldwide annually, and the number is expected to grow thanks to the very rapid increase of cell phone users in China. One survey predicts that the RF MEMS market could grow to over $1b in 2007. However, Ms. Waga mentioned that the handset product development cycle was very short (about 6 months, at least in Japan) and the system manufacturers were extremely sensitive about the price of components to be integrated in the handset unit. Replacement of passive components in the handsets by MEMS devices sounds very promising, but the replacement process would be very challenging for new entrants especially when innovation takes place rapidly in the field of proven technologies. Unless MEMS-based solutions prove extremely cost effective and highly reliable, the replacement scenario for handset applications would not be as rosy as it promised to be. Therefore, it is expected that RF MEMS devices would be first deployed in other applications such as semiconductor testing equipment, radar systems, and defense systems.

Lastly, the importance of packaging for successful commercialization of MEMS/MST devices was discussed. In this area, it is believed that Asia could play a significant role, since electronic device manufacturers in Japan, Korea, Taiwan and Singapore have developed advanced microelectronics packaging solutions which could be utilized for MEMS packaging. In addition, many companies in Asia have offered cheap labor for labor-intensive microelectronics (and increasingly MEMS) packaging services. Such expertise and solutions are expected to reduce the manufacturing costs associated with bringing MEMS/MST devices to the marketplace.