A Data-Driven Approach for Accurate Estimation and Visualization of Energy Savings from Advanced Lighting Controls

Despite occupancy-based switching and daylight-based dimming controls being widely believed to have tremendous energy saving potential, there is often a lot of variability in the actual savings across customer sites. A major challenge in a reliable, site-specific assessment of these advanced lighting controls is the skew associated with time-logging using a low-power clock. We develop a robust analytical approach based on grid-search optimization and linear regression to correct the clock skew by exploiting the information stored in the cyclical nature of occupancy patterns in commercial buildings. We provide independent validation of the results using illuminance data to illustrate the strength of our approach. We also conduct comprehensive sensitivity analyses of the results by varying the assumptions about the underlying parameters values. Our results demonstrate that believable visualizations and reliable savings estimates can be generated using a low-power clock and a set of data-driven algorithms and analytics.