Sensing in the Ubiqitous Computing Environment

The three papers Reactive Environments (Cooperstock et al., 1997), Augmented Reality (Buxton, 1997), and Proxemic Interaction (Ballendat et al, 2010) are investigating interactions by sensing the physical environment as an input method, which is an area of research within ubiquitous computing. Buxton identified that there is an input relationship between location, function, and distance, but only the location-function relationship is explored in Buxton’s conference room and Cooperstock’s meeting room. Once the user enters a designated location, there is some automated function that reacts to the user’s presence, for example, to configure the setup to a remote location. Proxemic Interaction extends both these pieces of work by exploring the function-distance and location-distance relationships. As an example location-distance relationship, if I walk closer to the TV, I am more likely to want to interact with it. As an example of a function-distance relationship, I am standing near the TV but I am holding another device (e.g. phone), I want to use the device and TV together.

These papers did not explicate that there a temporal relationship in addition to location, function, and distance. If I want to talk to someone, I will wait in front of his door until he is free. If I want to have a casual chat, then I’ll saunter by casually; I wouldn’t stay there if I saw that she was busy. Using Cooperstock’s scenario, if I place something on a projector, I am more likely to use the overhead projector if I stand there than if I walk away afterwards. (Perhaps I want to prepare the overhead for later use.) Ballendat’s paper alludes to this conclusion when referring to semifixed features like books and chairs, which can be moved around by the user, but they do not consider interactions that could change within the seconds scale.

There are some other points for discussion, which I won’t go into detail. The Reactive Systems paper addresses sensing-based interaction from requirements whereas the Proxemic paper deals with interactions that emerge given a specific set of technologies. There is significant research that needs to be done to identify the proper type of décor and feedback for sensing-based input methods (e.g., Miaosen and the fire door example). Lastly, they do not consider social implications such as privacy issues or imperfect systems such as noisy sensing in the environment.