Proxemic interactions in ubiquitous computing ecologies

Abstract

In this dissertation, I explore how the knowledge of people's and devices' spatial relationships -called proxemics -can be applied to the design of ubiquitous computing (ubicomp) interactions. Edward Hall's proxemics theory describes how people use spatial relationships -such as varying their distance or orientation -to mediate their interactions with other people around them. But in spite of the opportunities presented by people's natural understanding of proxemics, only a relatively small number of ubicomp installations incorporate proxemic information within interaction design.Therefore, my goal in this dissertation research is to inform the design of future proxemic-aware devices that -similar to people's natural expectations and use of proxemics -allow increasing connectivity and interaction possibilities when in proximity to people, other devices, or objects. Towards this goal, I explore how the finegrained knowledge of proxemic relationships between the entities in small-space ubicomp ecologies can be exploited in interaction design. In particular, I provide the following three major contributions: First, I operationalize proxemics for ubicomp interaction with the Proxemic Interactions framework that serves to guide the design of ubicomp applications. The framework describes how designers can consider fine-grained proxemic information to mediate people's interactions with digital devices, such as large digital surfaces or portable personal devices. I identify five key dimensions of proxemic measures (distance, orientation, movement, identity, and location) to consider when designing proxemicaware ubicomp systems. I also identify the gradual engagement design pattern as one particular strategy that allows designing system interactions that move from awareness, to reveal, to interaction.-iiiSecond, I design the Proximity Toolkit allowing ubicomp developers to rapidly prototype proxemic-aware ubicomp systems. The toolkit simplifies the development process by supplying higher-level information about proxemic relationships between the entities in ubicomp ecologies through an event-driven API and visual inspection tools.Third, I explore the design of three case studies of proxemic-aware systems that react continuously to people's and devices' proxemic relationships. The case studies explore the application of proxemics in small-space ubicomp ecologies by considering first person-to-device, then device-to-device, and finally person-to-person & device-to-device proxemic relationships. Together, they validate the toolkit's versatility and the application of the Proxemic Interactions framework.-iv -

Collocated Multi-user Gestural Interactions with Unmodified Wearable Devices

“…This also requires an instrumented environment; proximity is detected by an infrared-based motion capture system. The Proximity Toolkit [29] supplies fine-grained proxemic information between people and digital devices. The toolkit gathers data from various hardware sensors and transforms it into rich high-level proxemic information but is not targeting off-the-shelf devices.…”

EXCITE: EXploring Collaborative Interaction in Tracked Environments

“…Panoramic is an evaluation tool for ubicomp applications, providing historical sensor data visualizations [11]. Finally, the Proximity Toolkit allows recording and playback of tracked interaction sequences of multi-entity relationships (multiple people and devices) [5].…”

“…Internally EXCITE connects to the Proximity Toolkit [5] for capturing the tracking information of people's and devices position and orientation. The Proximity Toolkit uses depth-sensing cameras to track people's location, and a high end motion capturing system with infrared-reflective markers to track devices.…”

“…1). EXCITE leverages proxemic information-such as people's and devices' distance and orientation, captured with the Proximity Toolkit [5]-simultaneously with video data and event data from the tools, and allows video analysts to generate queries on the data as a whole. These queries are annotated on a timeline, allowing the ana lyst to scrub through synchronis ed video capture feeds to validate and further annotate study sessions.…”

Towards Intelligent Migration of User Interfaces

“…room or building), position (close to a display), movement (coming towards a display) and orientation (facing towards the display). These factors allow better contextual knowledge about the current situation, so that the system can determine, for instance, whether the public display should facilitate interaction with someone or just display information [1,15].…”