Microelectromechanical system (MEMS)-based components, such as sensors and actuators, began penetrating the wearable products market about a decade ago, when the first accelerometers replaced mechanical springs in pedometers and step counters.
While inertial sensors (accelerometers, gyroscopes, and magnetometers) and environmental sensors (pressure, humidity, and UV index devices) are widely used in consumer wearable products, proprietary technologies require custom MEMS sensors.
One adapted device, for example, is a MEMS-inspired glaucoma lens. The soft disposable silicone eyewear embeds a MEMS sensor that captures spontaneous circumferential changes at the corneoscleral area  . The ability to non-invasively measure the changes over a 24-hour period may be one of the most useful methods available to distinguish normal-tension eyes from glaucoma eyes. 
MEMS micro-actuators can also trigger micro-mirrors, at high frequencies, to reflect light to be projected onto a surface. The micro-mirrors enable a range of wearable products, including heads-up displays (HUDs), head-mounted displays (HMDs), and virtual reality (VR) goggles.
Micro-actuators also support wearable micro-pumps used to deliver insulin and other drugs. Debiotech, a medical device company based in Switzerland, for example, has developed a technology based on MEMS micro-actuators (see Figure 1). The patch-pump consists of a disposable unit and a control unit, which includes a micro-actuator pump engine  .
Let’s take a closer look at the role and future of MEMS-based sensors in wearable products.
Evolution of MEMS Sensors in Wearable Products
There are four major factors contributing to the adaptation and success of MEMS sensors in wearable products.
The size of MEMS sensors has been reduced significantly over the last few years. A decade ago, the smallest MEMS accelerometer required a real estate of about 35 mm2 on a printed circuit board (PCB). The latest high-performance accelerometers, however, require a space of only 4mm 2 [4, 5] — almost the size of a tomato seed. The tiny sensors contribute to the small form factor of wearable products, and therefore, to a greater acceptance of wearable products in various markets.
All wearable product manufacturers seek longer operation times for their battery-operated devices. MEMS suppliers have made great efforts to reduce the current consumption of their sensors, and they have achieved considerable success.
Just a few years ago, the current consumption of an accelerometer was approximately 0.7mA, and that of a stand-alone gyroscope was about 6mA. These days, a high-performance MEMS inertial module that consists of an accelerometer and a gyroscope consumes less than 1mA under similar sensor configurations  . The significant current consumption reduction has accelerated the adaptation of multiple sensors in wearables, and consequently, led to increased growth in product quantity.
The price of MEMS sensors over the last few years has dropped dramatically, making wearable products more affordable to average consumers. The price of an accelerometer about a decade ago was more than three dollars. Today, an accelerometer with much better performance and advanced features costs around thirty cents. The price of MEMS sensors is expected to continue to drop as the technology improves and more MEMS devices achieve the economies of scale.
Integration and Embedded Functionalities
The integration of multiple MEMS sensors and features in a single package has made the implementation of sensors in wearable products a much faster and easier task for hardware and software engineers. In addition to hardware integration, higher functionalities and embedded algorithms have offered application developers a powerful tool to speed up the completion of their projects.
Wearable products already on the market provide users a wide range of advanced functions and features.
Contextual awareness is one of the applications that takes great advantage of sensors’ data in a wearable product. Detecting human activity, an important building block of the contextual awareness, can be challenging and complex because of the different scenarios to be considered:
- A traveling user could be on foot, riding a bicycle, or driving a vehicle.
- A person on foot could be stationary, walking, running, or moving uphill or downhill. One’s actions could include a variety of activities, from swimming to jumping rope to playing tennis or basketball.
- A wearer has a variety of user-defined gestures to be considered: a glance, pickup, tap, swipe, or rotation, for example.
In addition to monitoring physical activity and behavior for general fitness purposes, context-aware systems embedded in wearable devices also play a major role in healthcare:
- Wearables for the elderly can be designed to detect potentially dangerous situations, for example, if a person has fallen.
- A wearable monitoring system detects changes or unusual patterns in a person’s activities that can be used to determine the early symptoms of a disease. The device can help to identify age-related diseases or severe medical conditions before they occur, including early symptoms of arthritis and Alzheimer’s disease.
- Wearable context-aware reminders can assist the elderly or people with disabilities in performing activities — such as making emergency calls via voice command and turning on lights when doors are opened.
All the scenarios we've discussed so far indicate the high degree of complexity that needs to be addressed in a wearable product with multiple sensors. Figure 2 shows some of the latest sensors that can be embedded into wearable products. Each of the sensors provides very useful data about the user, and her/his social and physical environments. The sensor data is fused together, processed, analyzed, and used to create valuable knowledge that plays a crucial role for making decisions.
Sensor suppliers have quickly learned that solely offering hardware is no longer enough to expand their share in today’s wearable market. In addition to hardware offerings, software is becoming a critical component for differentiation. Software features like “sensor fusion,” a capability that gathers and interprets data from several sources, can be used as building blocks for advanced applications and a development platform  .
An example of such a solution, called SensorTile, is illustrated in Figure 3. The small-form-factor printed circuit board is designed for wearable applications. Equipped with multiple sensors and a Cortex-M4 microcontroller, developers can assemble this board together with a cradle and an expansion board to build a complete development kit  .