Large networks of ultra-low resolution sensors in buildings
Christopher R. Wren · 2006
The occupants of a building generate patterns as they move from place to place, stand at a corner talking, or loiter by the coffee machine. These patterns leave their mark on every object in a building. Even a lowly carpet will eventually be able to tell you something about these patterns by how it wears. However, our automated systems are largely blind to these patterns: elevator, heating and cooling, lighting, information, safety, and security systems all depend on humans to translate these patterns into action. A cheap network of sensors can sense these patterns and provide useful information to context sensitive systems in a building. This paper reviews some of our work on systems that adapt to the patterns that people create in a building. Specifically we will discuss automatic geometric calibration of indoor sensor networks, and light-weight discovery of behavior patterns. We also present some new experiments that illustrate the importance of coarse, global information for understanding human behavior on a building-wide scale.