The paper has been accepted to ACM Buildsys 2017.
Indoor climate monitoring is crucial for building agility, enabling indoor environmental quality (IEQ) assessment, occupancy-based climate control, and context-aware services. While instrumenting the space with low-cost sensors like temperature, humidity, and light level, is a viable option, it is not scalable for expensive sensors like carbon dioxide (CO), ozone, particulate matters (in particular PM2.5), Volatile Organic Compounds (VOC), sound pressure level, which are standard parameters in IAQ assessment. Previously, these instruments are placed on a cart manually navigated throughout the space to take measurements, which can be inhibitively laborious.
To enable real-time scalable indoor monitoring, the study proposes to leverage autonomous sensing using a navigation-capable robot. we propose a ST interpolation method based on ST binning, and hierarchical estimation of global and local trends. A use case of ventilation efficiency is described, where autonomous mobile sensing is able to distinguish spatial patterns of air age from different ventilation settings. This information is useful to improve the indoor environmental quality and energy efficiency, and importantly, occupant health, well-being and comfort.