Locomotion@location

Abstract

Contextual information of persons can be comprised of a variety of different fragments. The sensor-based recognition of activities, which is one very important part of contextual information, is very well evaluated in laboratory surroundings with different sensor configurations. This paper presents the utilization of locomotion and location information inferred from sensor-readings in a real-world setting by applying a system that operates in an opportunistic and unobtrusive way. We let the rubber hit the road by exploiting locomotion and (in-door) location information in private households to optimize the energy consumption in terms of autonomous and implicit control of electronic appliances. By using onbody and environmental sensor devices, that are not presumably fixed, thus are accessed in an opportunistic manner, our system is able to safely control devices in terms of implicitly optimizing the energy consumption. We have conducted a field study in 15 households, where we have used the location and locomotion information of the residents to decide with a rule-based background intelligence, which electronic appliances can be safely turned off.