The advent of laptops in the 1990s, coinciding with the expansion of the Internet, freed us from the tethers of power cords and Ethernet cables. Then the explosive growth of cell phones and smartphones brought us unprecedented mobility and wireless connectivity. Today’s Wrist-top Revolution, coupled with the meteoric rise of the Internet of Things (IoT), is taking mobility to a whole new level: wearable computing.
In this article, we’ll examine the concepts behind the user-experience-driven design methodologies that are being used to create some of the most successful wearable products on the market. We’ll also consider the features and functions that drive a wearable product’s energy budget and computational requirements, including the selection of microcontrollers (MCUs) that meet the product’s design requirements.
New realities of the wrist-top revolution
Smart watches, activity trackers, wearable GPS devices, heart rate monitors and smart glasses are prime examples of the wearable products that generated an estimated $8 billion in global sales in 2013, according to Futuresource Consulting. Offering novel combinations of sophisticated functionality, easy-to-use connectivity, compact form factors, ultra-low-power processing, and wireless connectivity, wearable devices are giving rise to entirely new classes of personal electronics that help us stay healthier, better informed, and better equipped.
Although several leading smartphone manufacturers began experimenting with bulky wrist-top versions of their existing handset products several years ago, the wrist-top revolution kicked into high gear in early 2012 when innovative upstarts like the Pebble Smartwatch leapfrogged the smartphone makers with a new class of lightweight wrist-top devices that made it easier for end users to leverage the smartphones they already owned. Garmin, Samsung, Sony, Fitbit, Magellan (Figure 1 ), and other consumer electronics makers also joined the wrist-top revolution with their own smartwatches, activity trackers, and other wearable products.
Figure 1: The Magellan Echo smart sports watch leverages Silicon Labs’ EFM32 Gecko MCU to extend battery life up to 11 months using a single CR2032 coin-cell battery.
This market environment has encouraged the emergence of small, agile startups whose innovative products, such as the Misfit Shine fitness tracker (Figure 2 ), are successfully competing for market share with established players.
Figure 2: The Misfit Shine is an elegantly designed fitness tracker that achieves exceptional energy efficiency and long battery life with the EFM32 Gecko MCU.
A successful wearable device must deliver the right combination of price, performance, functionality, and battery life, as well as a unique look, feel, and behavior to differentiate itself from its competitors. MCUs, sensors, wireless electronics, and attractive user interfaces must be shoehorned into a small footprint that can be comfortably worn on the wrist or elsewhere on one’s body. Since such form-factor constraints leave little room for a battery, wearable systems must be extremely energy-efficient to achieve the longest possible operating periods between battery replacements or charges.
User experience drives winning designs
Integrating these diverse elements into a market-winning product requires complex design trade-offs to balance power, performance, functionality, and form factor. Several manufacturers have successfully navigated this new territory using a so-called “user experience-driven” design methodology that inverts many of the conventional priorities and practices used by embedded developers.
The design process for an embedded system typically begins with defining the functions and capabilities that will serve as the project’s top-level drivers. Conversely, designing a wearable product frequently begins with defining the ‘user experience it’ will need to produce. These requirements define a product by the way it looks, feels, and interacts with the end user, as well as the impressions, feelings, and emotions it evokes. The next step in this design process is to translate the user experience into a ‘use case’, a set of top-level functional requirements used to define the product’s hardware and software elements.
Apple was one of the early pioneers of this strategy. They used it with great success to define new markets and capture existing ones. If you have any doubts about the importance of a well-crafted user experience, consider how the Apple iPod’s unique control wheel, jewel-like case designs, and easy-to-use iTunes software helped the company transform and eventually dominate the digital music player market.
Defining the user experience
The requirements that define a wearable product’s user experience fall into two categories:
- Functionality – the unique look, feel, features and functions that differentiate a wearable product.
- Ease of use – a set of requirements that enables easy set-up, intuitive operation and minimal maintenance. Long battery life plays an important role in ease of use since having to recharge a wearable device every few days can be frustrating and cause users to abandon the product.
Together, these elements define a user experience that can be translated easily into a use case forming the foundation of a product’s design. Depending on the application, defining the user experience might involve designing a wearable case that has an inviting texture, ergonomic shape, and design elements that convey a specific feeling. Other products might require creation of special visual paradigms for controls and displays that make complex operations simple and intuitive.
Defining the use case
Once a product’s user experience has been clearly defined, it must then be translated into a use case whose functional requirements will drive the wearable product’s design. A detailed use case can provide important information that makes it easier to perform accurate trade-off studies for nearly every aspect of a wearable design.
A use case should include the tasks the wearable device is expected to perform, the required resources, and the conditions in which it is expected to operate. These details typically include the types of data the device will collect, how it will interact with users and other devices, anticipated operating environment (temperature, water resistance, impact resistance, etc.), operational modes (data collection and analysis, user interactions, communication, etc.), and how frequently it synchronizes with other devices.
Armed with these guidelines, the design team can start to identify the sensing, computing, and communication components that best meet the application’s requirements. Meanwhile, the bill of materials (BOM) cost and energy budgets are developed in parallel with the preliminary design requirements, giving the team the necessary parameters to converge on an optimal design approach.
Use case aids energy management
Because battery life playssuch an important role in wearable designs, let’s take a closer look atthe energy-related portion of the use-case-driven design process.
Toaccurately model how design choices affect a wearable device’s batterylife, the use case should include detailed descriptions of factorsaffecting energy consumption, such as:
- The type of data the device must collect from the outside world and how frequently it must be collected.
- Whether the user interacts with the device via an app, touch display, buttons, or a combination of these. If so, what types of information does it communicate, and how frequently will it be used?
- How the device communicates with other wearables, a smart phone, a local network, or the Internet. Power requirements can vary widely depending on the wireless interface (e.g. Bluetooth, Wi-Fi, or ZigBee) and how it is implemented.
- How often the device synchs or exchanges data with its peers or a host system. (Synching too frequently with a host system such as a smartphone can decrease battery life.)
Onceassembled, the use case should provide a detailed picture of thesystem’s various operating modes and how much time the device will spendin each one. This will be the foundation of a system energy budget andinform any design tradeoffs needed to maximize battery life.
Use case aids MCU selection, optimization
Theenergy-related portion of the use case should include as muchinformation as possible about the sensing, control, and computationaltasks the wearable device will be performing, as well as which of thesetasks will be performed by the MCU or one of its peripherals. This willaid in selecting the MCU family that best meets the wearableapplication’s requirements and developing strategies that make the mostof the MCU’s energy-friendly features.
You can create a goodinitial estimate or your wearable application’s computationalrequirements by identifying the software functions and algorithms itwill have to perform and how frequently they will occur. Consider, forexample, a fitness monitor in which the MCU senses the user’s physicalactivity through a multi-axis accelerometer, monitors cardiac activityusing an IR proximity sensor, and uses additional sensors to detecttemperature, humidity, blood oxygen level, and even ultraviolet (UV)light exposure (Figure 3 ). The MCU must then filter the rawsensor data to eliminate noise and other artifacts before determiningthe actual step count and frequency, or combine it with heart rate datato differentiate between specific types of activities and otherbiometric inputs.
Figure3: Health and fitness trackers and other wearables contain sensors thatmeasure physical activity and other biometrics such as UV lightexposure.
Among the several excellent 32-bitprocessor architectures used in modern MCUs, the ARM Cortex family of32-bit RISC CPUs has emerged as the leading processing core for embeddeddesigns thanks to its efficient architecture, easily-extensibleinstruction set, and large base of development tools and code libraries.
Over the years, ARM has created several series of its CortexCPU, each optimized around a specific set of requirements. The ARMCortex-M series of processor cores, for example, which was developedspecifically for deeply-embedded MCUs, is used in applications whereperformance must be balanced against energy efficiency and low solutioncost. The Cortex-M series provides core options that address a widerange of wearable design attributes including price, battery life,processing requirements, and type of display (Table 1 ).
Withinthe Cortex-M series, the M3 and M0+ cores are optimized forcost-sensitive applications that still require high-performancecomputing and fast system response to real-world events with low dynamicand static power consumption. The more complex and capable M4 coreoffers dramatically faster completion of the computation-intensivealgorithms frequently used in bio-monitoring applications. Its enhancedinstruction set includes a library of powerful digital signal processing(DSP) functions. The M4 core’s single-precision floating point unit(FPU) can dramatically shorten execution times, reducing the period theMCU must remain awake and thereby minimizing its overall energyfootprint.
Sleep deep for long life
To reduce theMCU’s impact on the wearable platform’s energy budget, it is importantto minimize the frequency and duration of any task that requires it toawaken from a low-power sleep mode. As such, the use case should includethe frequency with which the MCU’s various tasks are expected to occurand whether their execution is event- or schedule-driven.
One ofthe primary ways to optimize a low-power embedded design is to find thelowest sleep mode that still provides adequate response to real-timeevents. Most MCUs using the Cortex-M processing core support multiplesleep modes.
Silicon Labs’ EFM32 Gecko family, for example, usesstandard 32-bit ARM Cortex-M cores combined with an energy-optimizedperipheral set and clocking architecture. Designed specifically forenergy-sensitive applications, this architecture features a range ofpower modes enabling developers to achieve the optimal energy efficiencyrequired by wearables. Typical low-power modes used in many MCU architectures include:
- Sleep/Standby – Enables quick return to active mode (usually via interrupt) at the expense of slightly higher power consumption.
- Deep Sleep – Leaves the MCU’s critical elements active while disabling high-frequency system clocks and other non-essential loads.
- Stop – A deeper version of Deep Sleep mode that enables further power savings while retaining limited autonomous peripheral activity and fast wakeup.
- Off – This ‘near-death’ state preserves the minimum compliment of functionality needed to trigger wakeup from an external stimulus.
Smart peripherals, equally smart designs
ManyMCUs are equipped with at least a few peripherals that perform routinetimekeeping, I/O, and housekeeping tasks while the CPU remains in one ofits low-power sleep modes. Some MCUs are also equipped with autonomousperipherals that perform multiple functions (e.g., counters/timers,ADCs, DACs, GPIOs, serial transceivers, etc.) without CPU intervention.For example, all of the on-chip peripherals supported by EFM32 Gecko MCUs can function autonomously and remain active in one or more of the device’s sleep modes.
Inaddition to the counter/timer, ADCs, DACs, GPIOs, and serialcommunication elements, many ultra-low-power, ARM-based MCUs offer thefollowing peripherals:
- A capacitive sense controller that senses touchpad contact and coordinates within an n-by-n grid with minimal or no CPU intervention.
- An LCD driver that can drive numeric LCD or TFT display via DMAs from memory without CPU intervention.
- Analog comparators that enable monitoring of threshold voltages for alert/alarm conditions without CPU intervention.
Insome ultra-low-power MCU architectures, the activities of peripheralfunctions, including serial communications, counters/timers, analog anddigital comparators, and higher-level I/Os, can be coordinated by aseparate low-power bus that enables events and signals from oneperipheral to be used as input signals or triggers by other peripherals.This bus architecture can ensure timing-critical operation with minimalCPU overhead as well as reduced software overhead, resulting inextremely low-energy wearable designs and long battery life.
Winning the Wrist-top Revolution
Designingwinning products for the wrist-top revolution requires a deepunderstanding of the new realities of wearable application requirementsand a fresh approach to integrating complex technologies andhigh-performance components into space-and power-constrained designs.Smart watches, portable fitness trackers, smart glasses, and otherwearable computing devices are changing everything we know aboutdesigning portable electronics.
Wearables are rewriting therules for design engineers who must seamlessly integrate sophisticatedsensing, computing, display, and wireless technologies into affordable,compelling, ultra-compact designs that can operate for months on asingle user-replaceable battery or other limited energy sources. Newwearable computing products are appearing on the market at anever-increasing pace, resetting our expectations for the end userexperience with each design innovation. And this Wrist-top Revolutionhas only begun.
Raman Sharma , Director Field Marketing, Americas, joined Silicon Labs through the acquisition of Energy Micro, where he served as the VP ofSales, Americas. Raman now leads the field marketing efforts for SiliconLabs’ 32-bit MCUs and wireless products. Raman was formerly the GlobalMedical Segment Manager with Freescale Semiconductor and has fifteenyears of experience in product design and technical sales with companiesincluding OKI Semiconductor, Xilinx, and Compaq Computer.