TrojanSense, a participatory sensing framework for occupant-aware management of thermal comfort in campus buildings. (February 2020)

• Record Type: Journal Article
• Authors: Konis, Kyle; Blessenohl, Simon; Kedia, Naman; Rahane, Vishal
• Is Part Of: Building and environment. Volume 169(2020)
• Publication Date: 2020-02
• Subjects: Participatory sensing -- Thermal comfort -- Personal comfort models -- Machine learning -- Sustainability; Buildings -- Environmental engineering -- Periodicals; Building -- Research -- Periodicals; Constructions -- Technique de l'environnement -- Périodiques; Electronic journals; 696
• DOI: 10.1016/j.buildenv.2019.106588
• Languages: English
• ISSNs: 0360-1323

Abstract

Abstract: This paper describes the development of TrojanSense, a participatory sensing framework developed to collect, analyze and report user assessments of thermal preference at the campus scale with room-level spatial resolution. TrojanSense was developed with the goal of supporting initiatives to improve campus community engagement on issues of occupant thermal comfort, energy efficiency, and equity in the environmental control of university buildings. TrojanSense is an open-loop system, where outcomes pairing thermal preference and concurrent indoor temperature measurements are analyzed and reported with room-level resolution on a Campus Thermal Preference Map to support building operators and to inform campus policies related to thermal comfort and energy efficiency. The TrojanSense framework integrates wireless Bluetooth Low Energy (BLE) proximity beacons embedded in campus buildings to sense temperature and to automatically solicit occupant feedback using proximity-based prompts. Data from an exploratory pilot study are analyzed to examine the applicability of the approach for developing predictive models of thermal preference and for placing existing space conditioning assumptions in context with occupant feedback. Results demonstrate the potential for TrojanSense to identify overcooling using objective and subjective measures, increase the accuracy of predictions of thermal preference, and provide greater insight into the impact of outdoor weather on thermal preference to inform future climate-responsive control strategies. Highlights: Buildings lack a mechanism for feedback on occupant thermal comfort. A scalable participatory sensing framework was developed and tested. Proximity sensing is used to automate subjective surveys with room-level resolution. Responses are analyzed and mapped on Campus Thermal Preference Map. Data-driven comfort models improve accuracy over a static PMV model.