Big Business With Small Devices
If you watch the news, you can’t help but notice how everything keeps getting smaller and smaller. Everybody is talking about the newest microelectronic products and developments.
Despite the headlines, it has been rather difficult for assembly equipment manufacturers to make sound business judgments based on published technology forecasts. Microassembly is definitely a trend we should monitor closely, but will it become a big business in the near future? What does the market for microassembly really look like? How much investment in assembly technology can we expect? What are the technical challenges and opportunities for equipment suppliers? Which technologies are driving microassembly? What can we learn from the historical development of other technologies, such as electronics assembly and semiconductor packaging? What should suppliers of assembly automation do to prepare for when the microsystems market really booms?
For the moment, the market for microassembly equipment can best be analyzed by looking at currently available microsystems, which are becoming more and more sophisticated. These products typically have both active and passive components. Many include microelectromechanical systems (MEMS).
Modern automobiles contain many microsystems, such as sensors for air bags, acceleration and tire pressure. In medicine, microsystems can be seen in biochips, nebulizers, needleless injectors and miniature motors. Microsystems can be found in telecommunications equipment in the form of connectors and optical MEMS.
According to a 2002 report by the Network of Excellence in Multifunctional Microsystems (Zurich, Switzerland), the worldwide market for microsystems will grow from $22 billion in 2000 to $68 billion in 2005. That’s an average annual growth rate of 20 percent. With that kind of growth, there should be a tremendous need for production equipment for microsystems, especially automated assembly equipment.
How big is the market for such equipment? Unfortunately, there are no published figures. Microassembly is still too young a technology for reliable data to exist. Nevertheless, we can at least try to estimate the business opportunities.
Let’s suppose that 3 percent of every dollar spent on microsystems goes toward investment in microassembly equipment. This would put the worldwide market for microassembly equipment in 2005 at approximately $1.78 billion (2 billion euros). Some 60 percent of this market has already been captured by established machine builders for products such as ink-jet printer heads and disk drive heads. That leaves approximately $712 million (800 million euros) open for new and emerging products.
Now, quite a few microsystems manufacturers still build their own production equipment, mainly to protect their proprietary technology. And, in the automated assembly systems business, suppliers are typically stronger in their home markets than overseas. Therefore, we estimate that only 10 percent of the total market for microassembly equipment will be open to independent European machine builders. Of course, that’s still $178 million (200 million euros) in equipment investment to compete for!
To date, most publications on microsystems technology have been concerned with the parts manufacturing process. But, assembly equipment manufacturers are more interested in how two or more microparts can be assembled into a microsystem and how this job can be automated in the most effective way. What will this equipment look like? And how much will it cost?
Microsystems must be assembled quickly and precisely. But, how precisely? How fast? Mainstream automated equipment for assembling microelectronics and other small devices can achieve a positional accuracy of 10 microns with a throughput of approximately 1,000 units per hour. Laboratory systems with high-precision robots can achieve a positional accuracy of 1 micron, but only at much lower throughput rates. To meet demand, automated assembly equipment for microsystems will likely aim for high throughput rates at reduced accuracy.
As a result, microsystems engineers will be challenged to simplify their designs and optimize their devices for automatic assembly. This will reduce the accuracy requirements for assembly machines.
Another challenge for high-volume production of microsystems will be parts feeding. Fortunately, a wide range of reliable feeding devices has been developed for the electronics industry, and these should be adaptable for feeding micromechanical parts. Of course, the better oriented the parts can be supplied, the easier it will be to feed them. Parts suppliers can adapt their production equipment to the available forms of parts presentation, if some form of standardization regarding size and shape of micromechanical components could be achieved.
In the world of small devices, the traditional joining technologies of the macro world, such as welding, screwdriving, clinching and press-fitting, have little use. Instead, adhesives will be the primary method of joining micro parts together. To dispense this adhesive, some technology can be taken from the electronics industry. But, some equipment will have to be adapted to the more complex geometries of microassembly.
Manufacturers of equipment for the semiconductor packaging industry will step early into the promising MEMS market. Their experience with bonding technologies will give them a clear advantage.
Besides adhesive bonding, microsystems manufacturers will be able to borrow other processes from electronic component production, such as punching, bending and injection molding. Lasers are widely used by component manufacturers for welding, marking and lead trimming, and I anticipate that lasers will be important to microsystems manufacturers, too.
In short, assembly machine builders will need quite a bit of expertise to adapt and miniaturize their equipment for microassembly.
Lessons From History
Is all this new to us, or can we learn from other sources of know-how?
Microassembly does have a historical precedent. If you look at the evolution of electronics assembly over the past 20 years, all of the requirements for microassembly regarding speed, technology and accuracy have been met by the technology for surface mount electronics assembly. The miniaturization and enormous cost reduction seen in today’s electronic devices are only possible thanks to the productivity gains brought about by the surface mount process.
The tremendous demand for surface mount equipment was the key to business success for manufacturers such as Siemens, Fuji, Panasonic, Philips and Universal—even though their stories are stuck a bit with the current downturn in the electronics business.
In that light, automated production of microsystems can become as lucrative as surface mount technology was, once the conditions for cost-effective volume production can be met.
So, where do we stand with microassembly today? Looking back, it took almost 20 years until the majority of through-hole technology was replaced by surface mount technology. Within another 20 years, the next generation of surface mount technology will replace today’s packages, but the assembly process will basically be the same. The goal is still to attach intelligent silicon to the device with the fewest steps in the smallest space.
This can be seen in the progress of flip-chip technology, where the chip or die is picked directly from a silicon wafer and placed on the substrate. This same process is necessary for MEMS production. Because many of the proven technologies for surface mount assembly can easily be adapted to MEMS production, we expect that production technology for microsystems will develop in half the time that it took for surface mount equipment.
The fast and robust placement machines of today can provide a good platform for high-speed microassembly, because they have been optimized for accuracy as well as speed. Visual recognition and correction mechanisms are part of their control systems. Only the geometries and presentation of micro components will have to be adapted to the machines’ capabilities. As a result, one of the keys to successful mass production of microsystems will be to design parts-feeding methods that are similar to existing methods for feeding surface mount devices. Then, existing equipment can be adapted to microassembly and brought into production quickly and efficiently.
The success of surface mount technology can be attributed to three major factors:
1. a high degree of standardization in size, shape and geometry of parts and equipment.
2. parts with simple geometries designed for assembly in only two dimensions.
3. a mainstream interconnection or bonding process.
Vacuum is the primary method of picking and transporting surface mount devices. Before that could happen, however, component surfaces had to be standardized to allow for vacuum picking. And, that standardization process took time. On average, establishing a new packaging standard in the electronics assembly industry takes about 5 years from development to mass production.
If standardization is the key to success for microsystems assembly, who should drive such standardization? Just as the electronics industry has been guided by the Surface Mount Equipment Manufacturers Association (SMEMA), manufacturers of equipment and supplies for microsystems assembly should sit together and develop standards appropriate for their industry. The same issues that SMEMA focused on—parts presentation, interconnection and machine interfaces—should be the focus of standardization for manufacturers of equipment and supplies for microsystems assembly.
I encourage organizations such as the German Engineering Federation and the European Factory Automation Council to respond to the needs of assembly equipment suppliers by promoting and driving the standardization process. Resources for developing standards are scarce, and a lot of time and money can be saved by not reinventing the wheel over and over again.
Microsystems and MEMS are not yet a big business for assembly equipment suppliers, and there are some years to go before they will be. In the meantime, we need to gather reliable data about the technologies and demands for producing these devices. And, we must learn from similar technologies and standardize components and processes. Surface mount technology can be a good paradigm. Suppliers with a background in this industry will have a clear advantage over manufacturers of equipment for mechanical assemblies.
Standardization is the only way to survive in this market, where cost pressures will require microsystems manufacturers to continually increase the productivity of their equipment.