Most of these devices are small and irregular in shape. Engineering these needs special skills and manufacturing these needs special equipment and processes. Designers need to follow special design practices for wearable designs to be realised as successful products. There are very few books available for designers to design wearable products. This article tries to help design engineers to get an overall view of the design and manufacturing process of wearable devices.
Since wearable devices are normally made in large numbers and are miniature in size, use of contract manufacturing is recommended. The focus is on the design process and what designers should look for when they design wearable devices to be made by contract manufacturers. While this list is not exhaustive, it gives a fairly good view of the complexities involved in the wearable device design.
Wearable devices are no different from classic embedded systems that we are all familiar with. Block diagram of a wearable device is shown in Fig. 1. It has following four blocks:
1. Power supply and management circuit
2. Core processor
3. Analogue interface [analogue front-end (AFE) covering sensor inputs and actuator outputs]
4. Communication interface (wireless and wireline)
What differentiates wearable devices from conventional embedded systems is the power management circuitry. Most wearable devices are battery-operated (either primary or chargeable) and need special circuitry to control power consumption.
Power management circuitry alone cannot optimise power consumption. It is the combination of the power management circuit and the device software in the system that results in optimal power consumption.
To a large extent we can say that, the processing power needed by wearable devices is very low, as most devices just collect data from sensors, store and send it to the central application for processing. The central application could be a smartphone or, in the case of the Internet of Things, cloud-mounted software. This is essentially done to ensure that the battery lasts long and data is processed in a central location by a more powerful processor.
Most integrated circuits (ICs) use very high-speed complementary metal-oxide semiconductor (CMOS) technology. In CMOS technology, power consumption of an IC goes up as the operating frequency of the processor goes up. So most wearable devices use low-speed controllers, and sections of the microcontrollers (MCUs) can be shut down or operated at a low speed when not in use.
We also need to remember that, some wearable devices that are used in medical applications (like insulin pumps or carry-on ECG probes) need to be designed in a controlled way, as these are approved by regulatory bodies like FDA/CE. Most device vendors do not recommend their devices for medical usage to avoid getting sued in case of any failure.
Wearable devices can be classified into the following four major categories:
Sports and fitness. Sports and fitness devices primarily track movement (using accelerometers) and, in some cases, heartbeat. While worn, these normally track motion and not body parameters.
Personal health monitoring. These devices track heartbeat, body heat, blood oxygen (SpO2) and ECG (single or 3-lead) signals and, in some cases, have special sensors. These devices normally track body parameters and therefore need to be highly reliable, as readings may be used for lifestyle adjustments.
Tracking and monitoring. These devices have GSM/GPS functions and are primarily used for tracking the wearer, mostly in geo-fencing applications.
Medical applications. These are high-end devices that are typically worn by patients who need controlled dosages, periodically. These devices are regulated and controlled by regulatory agencies. For this, manufacturers need to follow strict processes during development and manufacture. Importance of the rigid process is to ensure that these devices are traceable and, in case of problem, the root cause can be analysed up to the component level.
Most wearables can handle very low analogue signals and therefore have substantial analogue circuitry to process these. These are also real-time devices and synchronise their operations to the time or date for data collection/measurement. Their functionality can be represented by the flowchart shown in Fig. 2.
Essential design elements
Let us now take a look at some generic design elements that impact a product. These impact both design and manufacture of wearable devices. A product’s performance, its manufacturing and usage characteristics all depend on how well the design is done.
First, let us understand the life cycle of a wearable product so that it is easy to understand the issues involved and how these can be addressed. Fig. 3 shows the typical life cycle of a wearable device. A device essentially goes through four major phases as shown in the figure. While the design, manufacture and use phases are sequential, support phase is basically concurrent, addressing the manufacture and use phases. The reason for taking a life-cycle based approach is to show how designs need to address issues that crop up in subsequent stages.
Design phase. During this phase there are a few things that designers need to know for a successful wearable design. These are:
1. Almost all wearable devices have sensors that use a very low signal level. This necessitates that the designers be very strong in analogue and mixed signal (combination of analogue and digital signals) circuit designs.
2. Since most wearables are battery-operated, designers need to be conversant in low-voltage analogue designs as well as power-conversion techniques (like boost regulators).