The key element in the consumer appeal of wearable devices lies in their touch-sensing HMI or human-machine interface—it provides an intuitive and responsive way of interacting via sliders and touch buttons in these devices. Wearable devices include earbuds, smart glasses, and smartwatches with a small touchscreen.
An unimaginable competition exists in the market for such types of wearable devices, continually driving innovation. The two major features over which manufacturers typically battle for supremacy and which matter particularly to consumers are—run time between battery charges, and the form factor. Consumers typically demand a long run-time between charges, and they want a balance between convenience, comfort, and a plethora of features, along with a sleek and attractive design. This is a considerable challenge for the designers and manufacturers.
For instance, while the user can turn off almost all functions in a wearable device like a smartwatch for long periods between user activity, the touch-sensing HMI must always remain on. This is because the touch intentions of the user are randomly timed. They can touch-activate their device any time they want to—there is no pattern that allows the device to know in advance when the user is about to touch-activate it.
Therefore, the device must continuously scan to detect a touch for the entire time it is powered up, leading to power consumption by the HMI subsystem, even during the low-power mode. The HMI subsystem is, therefore, a substantial contributor to the total power consumed by the device. Reducing the power consumption of the touch system can result in a substantial increase in the run-time between charges of the device.
Most wearable devices use the touch-sensing HMI as a typical method for waking up from a sleep state. These devices generally conserve power by entering a low-power touch detect function that operates it in a deep sleep mode. In this mode, the scanning takes place at a low refresh rate suitable for detecting any kind of touch event. In some devices, the user may be required to press and hold a button or tap the screen momentarily to wake the device.
In such cases, the power consumption optimization and the amount of power saved significantly depends on how slow it is possible to refresh the sensor. Therefore, it is always a tradeoff between a quick response to user touch and power consumption by the device. Moreover, touch HMI systems are notorious for the substantial amount of power they consume.
Commercial touch-sensing devices typically use microcontrollers. Their architecture mostly has a CPU with volatile and non-volatile memory support, an AFE or analog front-end to interface the touch-sensing element, digital logic functions, and I/Os.
The scanning operation typically involves CPU operation for initializing the touch-sensing system, configuring the sensing element, scanning the sensor, and processing the results to determine if a touch event has occurred.
In low-power mode, the device consumes less power as the refresh rate of the system reduces. This leads to fewer scans occurring each second, only just enough to detect if a touch event has occurred.
