Smart homes or homes that smart?

The following is a summary of challenges for ubiquitous computing in the home that Edwards and Grinter first published in the Proceedings of Ubicomp 2001. The past few years have seen an explosion of interconnected technologies in the home. For example, Jupiter Research predicts that 28 million US households will have a home network by 2006. Devices on the home data network are typically connected to allow printer and file sharing, and to facilitate multiple broadband users. Further, paralleling this increase in the number of interconnected data-centric devices is a corresponding increase in the complexity of the home audio/visual "network." We use the term "network" here because many home entertainment systems consist of multiple components, typically connected through analog and digital cables, and controlled through a number of remotes often equal to the number of components. Increasingly, these data-centric and media-centric networks are beginning to overlap, as users want to share content across both types of devices. The complexity of the home network already presents great hurdles for current users, even including many technophilic "early adopters." These problems will only get worse as the number and types of interconnected devices in the home grow. The HCI community must address a number of challenges before these technologies will become, as Mark Weiser termed it, calming, rather than an infuriating morass of incompatibility and opaque functionality. These challenges are not purely technical. Rather, they raise cross-cutting issues in the technical, social, and design domains. For example, consider the tangle of cabling that lurks behind most home stereo installations. As confusing and opaque as this is, the move to wireless interconnec-tions between components has the potential to make usability even worse. When physical connections are present, you know what's connected to what (even if you may have difficult tracing a particular cable), and further , what can be connected to what. A system with physical connections is also relatively stable; bar kicking loose a connector, I know that the system will stay in the same configuration as when I last touched it. This is not the case when wireless technologies are used: connections are "invisible," interconnectivity can potentially change without any apparent indication of the change, and "debugging" a problem is not as simple as following a cable. Paradoxically, while it is the "invisi-bility" of wireless connections that make them appealing , it is this same invisibility that brings a new host of …