Enriching mobile interaction with garment-based wearable computing devices

Abstract

Wearable computing is on the brink of moving from research to mainstream. The first simple products, such as fitness wristbands and smart watches, hit the mass market and achieved considerable market penetration. However, the number and versatility of research prototypes in the field of wearable computing is far beyond the available devices on the market. Particularly, smart garments as a specific type of wearable computer, have high potential to change the way we interact with computing systems. Due to the proximity to the user`s body, smart garments allow to unobtrusively sense implicit and explicit user input. Smart garments are capable of sensing physiological information, detecting touch input, and recognizing the movement of the user. In this thesis, we explore how smart garments can enrich mobile interaction. Employing a user-centered design process, we demonstrate how different input and output modalities can enrich interaction capabilities of mobile devices such as mobile phones or smart watches. To understand the context of use, we chart the design space for mobile interaction through wearable devices. We focus on the device placement on the body as well as interaction modality. We use a probe-based research approach to systematically investigate the possible inputs and outputs for garment based wearable computing devices. We develop six different research probes showing how mobile interaction benefits from wearable computing devices and what requirements these devices pose for mobile operating systems. On the input side, we look at explicit input using touch and mid-air gestures as well as implicit input using physiological signals. Although touch input is well known from mobile devices, the limited screen real estate as well as the occlusion of the display by the input finger are challenges that can be overcome with touch-enabled garments. Additionally, mid-air gestures provide a more sophisticated and abstract form of input. We present a gesture elicitation study to address the special requirements of mobile interaction and present the resulting gesture set. As garments are worn, they allow different physiological signals to be sensed. We explore how we can leverage these physiological signals for implicit input. We conduct a study assessing physiological information by focusing on the workload of drivers in an automotive setting. We show that we can infer the driver´s workload using these physiological signals. Beside the input capabilities of garments, we explore how garments can be used as output. We present research probes covering the most important output modalities, namely visual, auditory, and haptic. We explore how low resolution displays can serve as a context display and how and where content should be placed on such a display. For auditory output, we investigate a novel authentication mechanism utilizing the closeness of wearable devices to the body. We show that by probing audio cues through the head of the user and re-recording them, user authentication is feasible. Last, we investigate EMS as a haptic feedback method. We show that by actuating the user`s body, an embodied form of haptic feedback can be achieved. From the aforementioned research probes, we distilled a set of design recommendations. These recommendations are grouped into interaction-based and technology-based recommendations and serve as a basis for designing novel ways of mobile interaction. We implement a system based on these recommendations. The system supports developers in integrating wearable sensors and actuators by providing an easy to use API for accessing these devices. In conclusion, this thesis broadens the understanding of how garment-based wearable computing devices can enrich mobile interaction. It outlines challenges and opportunities on an interaction and technological level. The unique characteristics of smart garments make them a promising technology for making the next step in mobile interaction.