There is a wide range of sensors being evaluated for wearable applications, with new design architectures emerging that are aiming to boost new applications, reduce the system size and weight, and extend the battery life.
Sensors in wearable designs can vary from a simple sensor in a sticking plaster that connects via a passive wireless connection to a hub device, all the way through smart watches to a portable pack bristling with different sensing and wireless technologies. All this is leading to different sensor architectures depending on the target application, which in turn is driving modular development systems that can cope with the wide range of sensors and implementations that may be required.
Accelerometers and digital compass sensors are being used for navigation as these wearable designs are mobile, but these devices are also increasingly being used as a hub to manage other sensors. Accelerometers such as Freescale’s Xtrinsic MMA9553
are not just being used for navigation, but also for gesture control. This allows new applications for the control of wearable systems such as smart watches without having to use small keyboards or voice input. Instead, key movements of the device can trigger events to take sensor measurements.
Figure 1: Freescale's Xtrinsic MMA9553 accelerometer has been specifically designed to act as a hub for other sensors in a wearable design.
By including a microcontroller alongside the sensor, a device can manage other sensors in the wearable design. This reduces the complexity of the hardware design by taking the computational load away from the central processor and reducing the need to have all the sensors connecting directly. This in turn reduces the power consumption of the overall system, extending the battery life.
The most immediate application in wearable designs is measuring the performance of the body with a range of new medical sensors being developed specifically with such wearable designs in mind. A growing public interest in healthy living is driving the emergence of activity monitors, with a number of devices already available.
Health monitoring devices with Body Area Networks (BANs) technology developed at IMEC Belgium and the Holst Center enable accurate, non-invasive monitoring with clinical-grade functionality. This is paving the way for more effective and economical healthcare.
Samsung is working with IMEC on a platform that can combine different health sensors that are used in new ways and also open up the data to the cloud. The Simband platform was developed at the Holst Research Center in Eindhoven using technology from its parent, IMEC. This has been launched as a reference platform for developers to use to combine new sensors alongside the existing wearable design.
Figure 2: The Simband developed by Samsung uses a new LED sensor technology from IMEC, Holst for health measurements.
The optical sensors developed by IMEC use an LED that generates a range of light frequencies. These are used to measure the oxygen in the blood at different layers in the skin, and give a reading of the pulse. Combining this with an ECG electric measurement gives the time for blood to arrive at the wrist, providing an estimate of blood pressure. The Simband measures 14 x 34 mm and uses a 1 GHz dual core ARM A7 28 nm single chip that combines the processor, Wi-Fi and Bluetooth links. It also uses a new cloud-based open software platform called the Samsung Architecture for Multimodal Interactions (SAMI). This collects data from a range of sources, aggregates it and displays it on the wearable device. It extends this by allowing the sensors to securely store data in the cloud regardless of the source's format or structure. SAMI will also allow data to be controlled by the individual generating it and not by third parties, so that personal health data can be better protected.
IMEC has also been collaborating with a number of companies on wearable technology. Along with Holst, it worked with Olympus to develop a new low power single channel electrocardiography (ECG) acquisition chip that can be implanted in the body. The analog feature extraction in the chip allows precise monitoring of the signal activity in a selected frequency band by a digital signal processor in a smart watch.
Figure 3: A low-power single-channel ECG sensor has been developed by IMEC and Olympus.
The new low-power ECG acquisition chip advances the state-of-the-art by consuming only 680 nA when all features are active, and also provides competitive performance, such as input SNR greater than 70 dB and a common mode rejection CMRR > 90 dB without any external passive components.
Temperature sensing is also a key element in a wearable design, and the MCP9700/9700A and MCP9701/9701A family of Linear Active Thermistors from Microchip Technology includes low-cost, low-power sensors with an accuracy of ±2°C from 0°C to +70°C for the MCP9700A/9701A and ±4°C from 0°C to +70°C for the MCP9700/9701, while typically consuming 6 µA of current. Unlike resistive sensors such as thermistors, the Linear Active Thermistor IC does not require an additional signal-conditioning circuit. This eliminates the overhead of a biasing circuit as the voltage output pin can be directly connected to the ADC input of a microcontroller, reducing cost and power consumption.
The MCP9700/9700A and MCP9701/9701A temperature coefficients are scaled to provide a 1°C/bit resolution for an 8-bit ADC with a reference voltage of 2.5 V and 5 V. This provides a low-cost solution for applications that require measurement of a relative change of temperature, either for the external temperature or for the temperature of the wearer. When measuring relative change in temperature from +25°C, a typical accuracy of ±1°C can be achieved.
The devices are also immune to the effects of parasitic capacitance and can drive large capacitive loads. This provides flexibility in the PCB layout, as the device can be remotely located from the microcontroller, a helpful capability when space is at a premium in a wearable design. Adding some capacitance at the output also helps the output transient response by reducing overshoots or undershoots. However, capacitive load is not required for sensor output stability.
Freescale has put together a range of sensors in its Tower development system
that can be used to prototype wearable systems with different technologies. The Tower controller is modular and expandable with modules that provide easy-to-use, reconfigurable hardware alongside interchangeable peripheral modules for communications, memory and graphical LCD to make customization easy. Open source hardware and standardized specifications promote the development of additional modules for added functionality and customization.
Alongside the Tower system, Freescale has designed an open-source, scalable reference platform that gives OEMs the building blocks they need to rapidly develop a wide range of wearable product designs from a common platform.
Figure 4: Freescale's Tower development system allows multiple sensor and control technologies to be combined and easily programmed to speed up the design of wearable systems.
Unlike other wearable solutions, the new platform is not limited to just one form factor or product category. The highly flexible, system-level design kit supports embedded wireless charging, incorporates processors and sensors within a hybrid architecture for scalability and flexibility, and comes with open-source software. The wearables reference platform (WaRP) is aimed at multiple vertical segments such as sports monitors, smart glasses, activity trackers, smart watches and healthcare/medical applications.
The platform is built on Freescale’s i.MX 6SoloLite ARM Cortex-A9
apps processor as the core processing unit, supports the Android OS, and integrates production-grade silicon, software and hardware. The hybrid architecture is optimized for a low bill of materials (BOM) and features Freescale’s Xtrinsic MMA9553
turn-key pedometer, FXOS8700
electronic compass and an ARM Cortex-M0+ Kinetis KL16
The MMA9553 combines accelerometer MEMS transducers, signal conditioning, data conversion and a 32-bit programmable microcontroller for an intelligent, high precision, motion-sensing platform able to manage multiple sensor inputs. The device can act as a hub for other sensors, making system-level decisions required for sophisticated applications such as gesture recognition, pedometer functionality, tilt compensation and calibration and activity monitoring.
The MMA955xL device is programmed and configured with the CodeWarrior Development Studio for Microcontrollers software, version 10.1 or later. This standard integrated design environment (IDE) enables rapid implementation of custom algorithms and features to match the application. Using the master I²C port, the MMA955xL device can manage secondary sensors, such as pressure sensors, magnetometers, or gyroscopes. This allows sensor initialization, calibration, data compensation, and computation functions to be off-loaded from the system application processor. The MMA955xL also acts as an intelligent sensing hub and a highly configurable decision engine. Total system power consumption is kept to a minimum because the application processor is powered down until it’s needed.
The hybrid architecture allows designers to address new opportunities as the market evolves, and to scale and customize their designs from both a hardware and software perspective to develop a single product or a portfolio of devices with a range of different sensors.
The increasing use of sensors in wearable designs is driving innovation both in the sensor technologies and the system architectures. Using programmable, variable LEDs to provide medical sensing, integrated with connection to the cloud, is driving a new generation of health monitoring applications and devices from global giants such as Samsung.
Using key sensors such as an accelerometer as the hub for other sensors is allowing system designers to minimize the load on the central processor, reducing size and cost and boosting the battery life. In this way, careful consideration of the sensor architectures is allowing a new generation of wearable devices and applications to be developed.