By Roger H. Grace, President, Roger Grace Associates
Currently, only a few large volume "killer" applications of microsystems technology (MST) based solutions exits. These include read/write heads (1.3B units), inkjet bprint heads (750M units), airbag accelerometers (90M units), manifold absolute pressure sensors (30M units), and disposable blood pressure sensors (20M units). The above figures are 2002 values.
It is interesting to note that the automotive market has two of these limited number of killer applications but more importantly, that many emerging automotive electronic systems applications are expected to contain a large number of recently-developed MST devices. Since their first introduction into vehicle electronic engine control systems in the mid 70's, automotive MST application opportunities have grown immeasurably.
Currently, over 70 potential applications exist for these devices . Many of these have found their initial application in high-end vehicles (BMW 7xx and Mercedes Benz 5xx) where the cost is less important, providing performance/convenience differentiation for automakers rather than in lower-end vehicles, where cost is a much more important issue. Unlike other market sectors that have been adversely affected by the recent economic downturn such as fiber optic communications, the automotive sector continues to be robust worldwide, with 2003 production projected at 55.2M units, and expected growth at approximately 1.6% compound annual growth rate (CAGR) from 2002-2007. This steady growth, compounded by the continuous introduction of new safety, convenience, and performance functions, is creating a market growth for MST expected to be in the 12-14% CAGR range. This paper will address the well-established "killer' applications as well as the short term, recently introduced, and currently under development opportunities. Table I provides a summary of these applications.
A silicon MST solution replaced a coil and magnet LVDT) pressure sensor. The new approach afforded the system integrator with a more reliable, smaller, and (potentially) lower cost solution. This device measures the pressure in the intake manifold and fed the information back to a computer whose programmed algorithm determined the optimum air/fuel mixture to minimize uncombusted hydrocarbons and maximize fuel economy. The majority of vehicles today have one of these devices as part of the electronic engine control (EEC) system. However, alternate solutions of directly measuring the airflow currently exist, some using MST (Bosch), as well as those using discrete solutions (Hitachi). Major players in the MAP market include Bosch, Delphi, Denso, Kavlico, and Motorola.
Airbag Accelerometers. . The earlier implementations of this device introduced in the mid 80's were truly electro-mechanical in nature. They consisted of either a ball and tube approach(Breed Automotive), where a ball bearing would roll down a highly accurately-machined tube and strike a contact at its end when a suitable acceleration was sensed, or a conductive metal cylinder which was attached to a beryllium copper spring made to roll up an inclined plane and make contact with a conductive terminal as a result of car deceleration. This approach was offered by TRW and was dubbed the "Rollarmite." Both of these devices were large, provided an on/off signal, and had a cost of approximately $20 U.S. in automotive volumes. These solutions were displaced starting in the early 90's with MSTs. Not only did the new solution provide a more reliable, smaller approach but since it operated in the analog domain, the resulting crash pulse could be measured and compared to stored crash signals to determine if in fact a crash was under way versus a gang of rambunctious teenagers banging on the vehicle bumper to set off the airbag. In addition, initial designs only required one of these devices, located in a module in the passenger compartment (single point sensing), at a cost of $5.00. Their introduction into safety systems significantly reduced system part cost as well as the weight, cost, and labor associated with installing cable systems to connect the up to five accelerometers that were required to effectively sense a crash. Today, virtually all vehicles are equipped with at least one of these devices, with some vehicles using as many as eight to monitor frontal as well as broadside and rear crash scenarios. Major producers of crash accelerometers include Analog Devices, Denso, Motorola, and Sensonor/Infineon.
As mentioned earlier , many new systems are currently under development or have recently been deployed in limited production applications. Chief among these include:
Wheel Speed Sensors - Currently, variable reluctance (VR) sensors are used in the wheels of many vehicles to sense their rotation/speed for antilock braking systems. These discrete wire wound and magnet solutions are being replaced by Hall Effect (HE) sensors and anisotropic magnet resistive ratio (AMR) solutions which embody MST. These new solutions provide higher performance/more reliable solutions; however, they are currently slightly more expensive than their VR predecessors.
Tire Pressure Sensors - Tire underinflation has posed a major problem in vehicle safety. Currently, two solutions are vying for domination of this US federally mandated measurement. Problems with tire underinflation create excessive heat and wear, increasing the likelihood of a blowout. A number of companies have offered solutions to this problem by proposing to directly measure tire internal pressure using MST. These silicon devices are not new to the market but have been available for over 30 years in various forms. This solution is being challenged by the non-direct method of measuring tire rotation of using wheel speed sensors that measure differential wheel speed as a function of tire inflation in conjunction with microcontroller algorithms. Although the United States National Traffic Highway Safety Administration (NHTSA) has not yet selected a specific approach to solving the underinflation problems, it has dictated an introduction schedule whereby all new vehicles that will operate on U.S. roads must be fitted with one of these systems no later than 2006. This certainly constitutes a killer application with an average of 4.2 tires/sensors per vehicle and over 16 million vehicles sold in the U.S. yearly. In addition, non-US applications are considered significant due to the federal highway safety organizations of many countries dictating similar procedures. The enhanced accuracy of the direct pressure measurement system is greatly offset by its excessive systems solution cost of $65 - 80 per vehicle, versus that of the wheel speed based solution. Target price for the module consisting of a pressure sensor, conditioning electronics, and a wireless transmitter package is being pegged at $12-15 (US) per wheel with a high volume price target at $5.00 per wheel by 2007. Schrader currently has approximately 60% of the total available market with for this application including Bosch, Motorola, and Siemens as upcoming viable competitors.
Rate/Yaw Sensors - These sensors measure the rate of rotation about a central axis which is most important in determining vehicle location when used in conjunction with a Global Positioning System (GPS), or to determine vehicle orientation as it begins to roll over or go into an uncontrolled skid. Early yaw rate sensors found their first application in the mid 1990's on the Mercedes Benz S500. The devices, designed and manufactured by British Aerospace, were developed from designs previously attributed to avionics guidance systems, cost approximately $80-100 (U.S.), and used piezoelectric sensing elements. Systron Donner brought the first quartz resonator MST rate gyro (gyrochip) to the automotive market in a high-end Cadillac platform for vehicle dynamic control (VDC) applications in the mid 1990's. These devices currently sell in the $20-25 range (US) and are installed in many U.S. and European vehicles. Recently, Analog Devices, Bosch ,Denso, and Motorola, and have introduced silicon MST to this application. It is expected that these devices will gain favor in many high (both in cost and in cross-section) vehicles over the next few years, especially as the number of SUVs continues to increase dramatically.
The technology associated with automotive vehicles and their MST sensors have come a long way in the two decades since the introduction of the first MST. Today, only two very high volume applications exist. As the need for safer, more fuel economic, and convenience-enriched vehicles becomes more significant, there is no doubt that many of the 70+ possible applications of MST devices will find their way into significant production levels for vehicles worldwide. A number of other significant high volume applications are expected to be realized in the next few years including high pressure diesel and non diesel fuel rail sensors and gasoline tank fuel evaporation sensors.
The major challenge to MST suppliers is to meet the enormous cost pressures imposed by the car manufacturers and the first-tier suppliers, the 100,000 mile and 10 year lifetimes of the parts and the length of time (typically four years) from prototype to production necessary to bring the product to market, and the necessity to produce tens of millions of these devices each year. These requirements eliminate all but a few of the large number of MST suppliers currently in business.
Roger H. Grace is President of Roger Grace Associates, a consultancy specializing in Microsystems Strategic Marketing activities and located in San Francisco, California, His is also the President of the Micro and Nanotechnology Commercialization Foundation (MANCEF). www.rgrace.com
 Roger H. Grace, "Automotive Applications of MEMS/MST: The Migration from Discrete Solutions." Advanced Microsystems Automotive Applications Proceedings, Berlin, Germany, April 2002 pp. 1-14.