COD: a dataset of commercial building occupancy traces

摘要: In this poster we introduce the Commercial Occupancy Dataset (COD), a high-resolution long-term dataset of occupancy traces in commercial office building spanning 9 months and covering zone-level for five different spaces containing more than 90,000 enter/exit events over time period. data contains rich spatial-temporal information about users their usage space facilities. However, obtaining accurate is very challenging task due to limitation existing sensing technologies. A novel depth-imaging based solution estimate counts was deployed doorways an generate dataset. The has high resolutions all three major dimensions: temporal, spatial state. This allows applications such as energy simulation, modeling human-in-the-loop HVAC control which enhance efficiency human comfort. Towards objective, provide case study demonstrate utility understanding occupancy.