Micro-Electro-Mechanical Systems (MEMS) are more prevalent than ever, especially in the popular configurations known today as “wearables.” Leading MEMS and sensors in the multitude of wearable technologies include accelerometers, gyros, magnetometers, microphones, UV sensors, glucose monitors, barometers, humidity detectors, and heart rate instruments.
Wearable systems include microcontrollers that detect, interpret, and communicate multiple signals. The combined MEMS, sensors, and hub electronics within wearables communicate via interconnects.
We traditionally think of electronics in a printed circuit board format housing various components connected by metal traces. Future categories of wearable devices may not fit this archetype, such as a MEMS and sensor array covering an entire jacket. Would the designers of such a jacket really want a bulky array of devices sewn into the fabric that require different batteries and wireless antennae for power and communications? How can we best connect arrays of MEMS that monitor and transfer data over a large range of surface areas?
The challenge of wearable interconnects is to combine sensor devices with miniaturized conductive grid systems that provide low signal loss and increased mechanical flexibility. The requirement for flexibility in wearable devices necessitates the development of novel technologies for device manufacturing and processing. As an example, new manufacturing methods include roll-to-roll technology, which produces bendable interconnect systems.
Some of the most popular signals to monitor in the world of wearables are movement, direction, sound, light, and pressure. BioMEMS and biosensors provide health data by monitoring heart rate, pulse rate, oxygen in blood, blood pressure, respiration, hydration, and chemistry of perspiration. Interconnects enable power and communication between sensor signals and hub electronics (such as microcontrollers).
Additionally, the devices bring wireless signals to host devices.
The market size of MEMS (Figures 1 and 2) and sensor systems for wearables grows daily. In fact, research firm IHS Technology predicts over 450 million MEMS and sensors shipments for wearables by 2019. In this article, we will review the MEMS devices in today's wearables, and demonstrate how a flexible interconnect capability can better support the increasingly popular MEMS-powered technologies.
One of most popular wearable technologies today is the fitness and health monitor. Pedometers, wristband trackers (Figure 3), and portable blood pressure monitors commonly contain accelerometers. Accelerometers are MEMS devices that detect acceleration, vibration, shock, and tilt. When coupled with MEMS gyroscopes, rotation can also be sensed, providing additional accuracy in movement. Current products include multi-axis accelerometers and multi-axis gyros in one package.
Most MEMS accelerometers today use a differential capacitive measurement system which calculates g force (acceleration). Upon experiencing an acceleration, a MEMS mass attached to a spring will move in distance — the moveable mass is one plate of a capacitor. An adjacent fixed mass is the other plate of the capacitor. The relative change in displacement between the two masses results in a change in capacitance. The signal is typically converted into volts subsequently sensed by a data acquisition system. A microprocessor then performs the proper interpretation of the data through algorithms that determine the output of the accelerometer.
A MEMS gyroscope measures rotation; the rotational speed is also known as angular velocity. The gyroscope therefore provides movement information in the absence of any linear acceleration. An on-axis gyro will measure a single axis rotation, while a three-axis gyro will measure rotation around all three axes. Since the gyro measures in units of degrees/second, small and slow movements are detectable with a MEMS gyroscope, through a small resonating MEMS mass that shifts as angular velocity changes. The output is amplified and fed into a microcontroller for interpretation.
Detection of the Earth's magnetic field using MEMS magnetometers has the benefit of not requiring metallic materials. The coupling of a MEMS magnetometer with an accelerometer in the wearable technology allows for tilt compensation, so that pitch and roll can be accurately calculated. The magnetic sensor requires calibration to convert raw data to a heading calculation (also known as yaw). Coupling a MEMS magnetometer with a MEMS accelerometer results in a 6-axis device; the combination uses approximately 50% less power and can boost magnetic signal by 30%.
Researchers are developing methods to use a magnetometer and the human body in wireless communication between two wearable devices. Gesture recognition, a promising feature for many MEMS devices, can be performed with magnetometer device technology.
The Interconnect Challenge
We have demonstrated above that multiple MEMS devices have been combined in one package for improved functionality. To enable light and small wearable sensor systems, sensor device sizes are being reduced using very small packaging. Compact packaging approaches include wafer-level, chip-scale packaging and 3D packaging with interposers or fan-out technology. The latter allows MEMS and microcontrollers to be stacked and arranged as one unit. Accommodation of small sensor packaging in wearable interconnects, with a form factor that moves and flexes with ease, is needed for further advancement of the technology.
Most wearable technologies today monitor human action. Unfortunately, today's rigid circuits threaten to hinder the ease and comfort of the consumer's movement. Wearable electronics worn in shirts, pants, and jackets will cover a wide surface area and must not be readily felt during the normal motions of the user. Typical human movements such as walking, running, bending, and twisting cannot be hindered by tight wiring designs — this is why interconnect flexibility will be most important in smart clothing applications.
To this end, let's identify today's less flexible technologies, followed by emerging manufacturing techniques for bendable electronics. The flexible manufacturing methods have the potential to disrupt the current barriers that limit wearables to predominantly watch-sized electronics seen in the present market.
The flexibility requirements of interconnect approaches are different for various wearable products. Some wearables employ small, rigid printed circuit boards (PCBs) to mount the tiny MEMS devices and associated hub electronics (Figure 4). Routing can be achieved using rigid-flexible printed circuits which are typically printed on sheets of polyimide. The additional flexibility of the serpentine geometries provides limited bending and twisting upon movement, especially compared to the typically rigid, resin-based PCBs.
“Flex to fit” devices are flexible circuits that extend, but don't move back to their original position. Such “flex” technologies support specific applications that require lengths of interconnect with specific geometric routing. Bendable electronics can provide an additional level of flexibility that the previously mentioned technologies cannot offer, as they can bend at higher angles and more orientations than printed circuits on polyimide.
Recently, bendable electronics have been manufactured using roll-to-roll nano-processing. Many variations on roll-to-roll techniques have been devised and show promise for some wearable product interconnect needs.
Roll-to-roll processes for nano-imprint lithography stretch a flexible substrate across a system of rollers, depositing material as the substrate rolls along the production line. Instead of consisting of individual silicon wafers, the substrate can be a continuous bendable surface analogous to a poster roll. Some strategies deposit photoresist that fills an imprint template and is hardened by a flash of UV light.
Other variations print inks containing nanoparticles. Roll-to-roll gravure printing takes processes familiar to the printing of newspapers and applies them to printed electronics. The roll-to-roll nano-manufacturing processes provide advantages in flexible substrates, including the ability to print over large surface areas, offer high-production throughput, and make multiple products with minimal disruption to equipment setup.
Although roll-to-roll processes do not typically achieve the same feature-size resolution as conventional lithography, progress is rapidly improving on this front. Roll-to-roll processes have demonstrable benefits for applications which are bottle-necked by the need for a large surface area versus a small feature size. The challenges in roll-to-roll printing involve bringing small-scale laboratory processes to higher roller speeds and wider surface areas, and mitigating the mechanical wear and tear on the manufacturing equipment to minimize downtime and cost. Precise alignment of multiple deposited layers, under high roller speeds, is also a crucial task for manufacturers.
Wearable clothing will need very flexible interconnects that call for clever approaches. Electronic sportswear today embeds devices into sewn small pockets in the garment. Embedding the electronics in the clothing itself is the goal. Thin bendable interconnects with reduced geometries will be incorporated into wearable garments. Developments in electronic inks and conductive yarns as part of the product fabric are interconnect technologies currently in research.
Reliability will also challenge the future of wearables. Clothing gets dirty, thrown around, and spilled on. Temperatures in use can range from dry subzero to wet high temperatures. Most clothing additionally will be washed from 10 to 100 times, subjecting the fabric and embedded electronics to an extreme range of humidities and temperatures. Any embedded system therefore will see conditions of use that are quite different from those typically experienced by household or office electrical systems.
One of the main definitions of a wearable is that it is always on. Thus, designing for low power consumption is critical, as well as detection of low activity (for example: fitness trackers) and the enabling of variable power settings. Designing the system to reduce power consumption, and using the lowest-power devices with negligible signal loss interconnects, will result in extending the time between battery charges.
Bendable and smaller batteries are in development, but capacity goals still must be attained. Thus, lowering power consumption through intelligent algorithms, choice of microcontrollers and hubs, coupling of MEMS devices in smaller device packaging, and low interconnect losses are essential power-related areas for research in wearables.
The Future of Wearables
There are a variety of opportunities for future work and research in wearables. Interconnect grids that connect the system without losing signal, or interfering with the form and fit of the wearable, is an exciting development area. Power improvements through smarter design and integration of MEMS devices, microcontroller development, smart software, variable power options, and increased battery capacity in small form factor are underway.
Current technologies are being employed in ways developers did not imagine. Researchers, for example, are using fitness tracker data to study sedentary behavior. Health monitoring will continue to develop, and will include more smart clothes, fabrics, and new devices. Solving the interconnect challenges in various wearable technologies will be essential to a smart and reliable system.
This article was written by Allyson Hartzell and Andrew Spann of Veryst Engineering, LLC (Needham, MA). For more information, visit Here .