We introduce a novel method for selecting and controlling smart appliances in physical spaces through a head-worn computing device with near-eye display and wireless communication. We augment a commercial wearable computing device, Google Glass, with a narrow-beam IR emitter for this purpose. This configuration yields a usable beam width of 2 to 4 feet (60 to 120cm) for targeting at room scale. We describe a disambiguation technique if infrared targeting hits multiple targets simultaneously. A target acquisition study with 14 participants shows that selection using head orientation with our device outperforms list selection on a wearable device. We also report qualitative data from using our device to control multiple appliances in a smart home scenario.

Figure 1. Left: Our head orientation-based selection techniques use an IR emitter -multiple targets may fall within its illumination area. Center: We offer two list-based refinement techniques -Naive IR uses alphabetical ordering; Intensity IR orders targets by IR intensity. Right: Using Head-motion Refinement technique, users can refine their selection through head orientation refinement in a quasi-mode when they hold the touchpad. ABSTRACTEmerging head-worn computing devices can enable interactions with smart objects in physical spaces. We present the iterative design and evaluation of HOBS -a Head OrientationBased Selection technique for interacting with these devices at a distance. We augment a commercial wearable device, Google Glass, with an infrared (IR) emitter to select targets equipped with IR receivers. Our first design shows that a naive IR implementation can outperform list selection, but has poor performance when refinement between multiple targets is needed. A second design uses IR intensity measurement at targets to improve refinement. To address the lack of natural mapping of on-screen target lists to spatial target location, our third design infers a spatial data structure of the targets enabling a natural head-motion based disambiguation. Finally, we demonstrate a universal remote control application using HOBS and report qualitative user impressions.