Think Outside the Chip: MEMS-Based Systems Solutions

MEMS is an acronym for Microelectro mechanical Systems; however, most MEMS implementations to date have not been systems at all, but rather devices. This article reports the constituents and some applications of what is defined as MEMS-based systems solutions, or MBSS. In Europe, this concept is commonly referred to as ”Smart Systems In tegration.” These MBSS use front-end MEMS devices — either one or a combination of many sensors, actuators, and/or structures — that work in conjunction with several other devices including signal conditioning commonly using application specific integrated circuits (ASICs), digital signal processing (DSP) with embedded microcontrollers and software, energy creation and storage, and networking communications functions.

All of these functions need to be interconnected and contained in a small, robust, low-cost package that has the ability to be tested in a highthroughput fashion. The concepts of classical system engineering of bringing the design team together from day one; co-design principles and modeling that acknowledge the interaction of the MEMS, other electronics, and packaging; reliability analysis; and embracing design for manufacturing and test are the principles that drive this approach (Figure 1). All of this is driven by the specific application.

Approaches to MEMS-Based Systems Solutions

Market research by Roger Grace Associates on MBSS applications has established that they fall into two categories. Category one has an “enabling engine” that drives the solution. This concept is driven by MEMS technology that is the intellectual property of the creator of the solution. As an example, a high-sensitivity accelerometer created by HP is the enabler for a wireless autonomous sensor network for seismic oil and gas exploration applications1 as well as the hand-held Phazer Near Infrared (NIR) spectrometer from Thermo Fisher¹.

The other approach, “commoditized integration,” uses standard off-the-shelf MEMS devices in conjunction with ASICs/DSPs with proprietary application software and proprietary packaging to create a solution. Examples of this approach are Schrader Electronics1 in its automotive tire pressure monitoring systems automobiles using off-the-shelf pressure and motion sensors; Hillcrest Laboratories¹ and Movea1 in their human-machine interface devices for gesture recognition using MEMS accelerometers and gyros; and Body Media for their weight-control/calorie-counting, bicep-wearable Link and Core system using a purchased three-axis MEMS accelerometer with internally developed sensors including (non-MEMS) skin temperature, galvanic skin response, and heat flux.

The value added to these “commoditized” solutions is the developers’ in-depth knowledge of the application, including the data acquisition, storage, processing, and transmission requirements, and the ability to successfully accomplish intelligent systems integration and provide software programming specific to the application. All of this is neatly packaged in a low-cost and robust fashion.

Applications

The applications for MBSS are wide and varied, as mentioned above. Based on our market research, one of the more significant opportunities for MBSS are autonomous wireless sensor networks (WASN). Many applications are currently under development at research universities and laboratories worldwide. One of major investigation is the use of WASN in infrastructure monitoring. Applications include civil engineering structures including bridges, dams, and buildings, as well as for structural monitoring of ships’ hulls and aircraft structures. The University of Michigan’s Center for Wireless Integrated MicroSensors and Systems (WIMSS) is part of a team working on a National Institute of Standards and Technology (NIST) grant to instrument bridges with accelerometers and strain sensors, collect the data, and wirelessly send the data to a central data collection platform to monitor the bridge system dynamic response in real time and over time under traffic conditions1. Similar systems are being developed jointly by MEMSIC and the University of Illinois Urbana Champaign¹.

In the United States alone, there are 583,000 bridges. The American Society of Civil Engineers (ASCE) reports that approximately 25% of these are structurally deficient or functionally obsolete. ASCE also reports that of the 76,926 dams that exist in the US, 1,819 have high hazard potential and are now considered deficient with the public at risk. Many of the over 2 million lines of natural gas pipelines in the US are considered antiquated and prone to failure. Finally, many of the over 100,000 miles of levees are reported to have structural integrity problems.

This broad and aging infrastructure is a prime candidate for wireless sensor networks. The University of Michigan WIMSS Center team has instrumented approximately ten bridges worldwide using this WASN approach. In keeping with the WIMS “mantra” of integration, they have adopted an approach using 3D chip stacking as a means to a robust and space-conscious package to their solution (Figure 2). Researchers believe that this real-time monitoring approach can be an early indicator of structural failure, resulting in the possible saving of lives in addition to reducing maintenance and inspection costs.

Future Application Opportunities

MBSS principles can be applied most effectively to many applications. Our research points to “smart building” systems and home-based point-of-care patient monitoring as holding promise for future large-volume applications of MBSS. In the area of smart buildings, the same principles discussed in the infrastructure monitoring above can be applied to make buildings more energy efficient, safer, and the environment more pleasant and healthy to its occupants. All of the constituents defined in Figure 1 apply here: sensors including temperature, airflow rate, humidity, air quality/gas constituents, light levels, motion sensors, and others are readily available as off-the-shelf commodities. Wireless network chips are also off-the-shelf commodities. Also readily available are ASICs and high-performance power sources (i.e. batteries). The challenge from an engineering perspective is to create the systems-based solution, the applications algorithms, and the integration of all of these functionalities into a small and robust package made for easy deployment in buildings at a price that provides users with quick paybacks.

Conclusions

Although MBSS has existed for many years, the proliferation of this approach is being driven from the tech nology push perspective with the availability of low-cost MEMS devices, signal conditioning ASICs, and DSP, as well as lowcost packaging such as wafer scale packaging and high-throughput testing. Fortunately, there is an “applications/ market” pull situation that co-exists, making this a very interesting situation for widespread adoption in numerous high-growth applications.

The need for gesture recognition in games, toys, computer peripherals (mice), medical, sports, and fitness bodes well for this approach. Add - itionally, as the need to better understand the quality of food and water, as well as the chemical and biological composition of many substances to enhance society’s quality of life, the MBSS approach for spectrometers and chromatographs for easy-to-use and low-cost handheld instruments will create many application successes.

However, the ultimate forcing factor to the adoption of MBSS by developers will be the continuing need for welldefined and defensible product differentiation and higher profit margins vis-à-vis higher levels of integration and value-added that optimally satisfy the customer’s applications needs.

This article was written by Roger H. Grace, President, Roger Grace Associates, Naples, FL. For more information, Click Here .

---