Abstract:

One prediction about this future of pervasive technology is that people will carry the tools needed to interface with technological resources sprinkled through out the environment. A problem with this vision is the dark side of the network effect: early adopters will end up carrying around interfaces for technology that largely does not yet exist, and building managers will question the value of installing technology with features that almost no one will be able to use. An intermediate solution is that certain buildings with specific needs for efficiency or security (such as hospitals) may become smart, with technology insinuated into particular spaces. Since many, or even most of the people in these spaces will not have the technology to interface directly with the new pervasive resources, we must think of the interaction idiom as initially being closer to the notion of smart environments. These environments will have to sense, interpret, and facilitate the actions of the inhabitants, possibly with very little help from technology attached to the people involved, or even their cooperation. We survey a body of work on perceptual tools for smart buildings, built on the sensor network model, and focused on the idea that statistical methods and population dynamics can provide valuable information even in situations where detection of individual instances of behavior may be difficult to detect. These are some of the tools which will fuel the building optimization applications that will justify the efforts of early adopters to build smart buildings studded with pervasive technology.