Automated Metadata Construction to Support Portable Building Applications

Abstract

Commercial buildings consume nearly 19% of delivered energy in the U.S, nearly half (42%) of which is consumed in buildings with digital control systems [23] comprised of wired sensor networks. These sensors have scant metadata, and are represented by "tags" which are obscure, buildingspecific and not machine parseable. We develop a human-inthe-loop synthesis technique which uses syntactic and datadriven steps to parse these sensor tags into a common namespace, which can enable portable building applications. We show that our technique allows an expert to fully parse a large fraction (70%) of the tags with 24, 15 and 43 examples for three large commercial buildings comprising 1586, 2522 and 1865 sensors respectively, and deploy three portable applications on two buildings with less than 30 examples.

Smart buildings as Cyber-Physical Systems: Data-driven predictive control strategies for energy efficiency

“…These fields may be complemented by approaches such [145,146] that analyze the meta-information in a BMS, e.g. the data point names as well as the data, by data mining techniques to identify and map variables of interest correctly with minimal human intervention.…”

FlashProfile: a framework for synthesizing data profiles

“…Although profiles generated by FlashProfile are primarily aimed at data understanding, in § 6 we show that they may aid PBE applications, such as Flash Fill (Gulwani 2011) for data transformation. Bhattacharya et al (2015) also utilize hierarchical clustering to group together sensors used in building automation based on their tags. However, they use a fixed set of domainspecific features for tags and do not learn a pattern-based profile.…”

AutoPlug: An automated metadata service for smart outlets

“…Further, the problem of automated metadata in the context of the commercial buildings was addressed in these recent works [16,11]. In Jungkun et al [16], authors develop a framework to automatically infer a sensor label based on its time-series data using off-the-shelf classifier.…”

“…In Jungkun et al [16], authors develop a framework to automatically infer a sensor label based on its time-series data using off-the-shelf classifier. Bhattacharya et al [11] present a syntactic and data-driven approach to parsing sensor names to common name spaces. These works target building automation/management systems and the commercial building space where metadata associations seldom change, on the other hand, AutoPlug targets home automation systems and the residential home space where metadata association in the context of smart outlets change frequently.…”