Going smart, staying confused: Perceptions and use of smart thermostats in British homes

Testing smarter control and feedback with users: Time, temperature and space in household heating preferences and practices in a Living Laboratory

“…Lastly, we fully acknowledge that many of our findings may not be necessarily exploratory, but are instead confirmatory, supporting findings from earlier studies of the variation in heating practices and preferences (e.g. Dimitrokali et al 2015;Eon et al 2017;Kane et al 2015;Miu et al 2019), a fact that we believe still has value for its ability to triangulate earlier research.…”

“…For example, an Australian study which monitored ten households by Eon et al (2015) found variation in heating practices according to family size and employment status, with higher levels of heating linked to the amount of time people spent at home (those working out of the home heated their homes only in the mornings and evenings, whilst stay-at-home parents with kids used their heating all day). To the contrary, in a UK-based study with 50 participants, Miu et al (2019) found that despite participants having fairly regular occupancy patterns, they did not in fact match their heating controls to their routines but left the heating on even when they were not at home. The notion of flexibility in energy service practice has been highlighted for those energy services that use electricity (e.g.…”

Smart energy systems beyond the age of COVID-19: Towards a new order of monitoring, disciplining and sanctioning energy behavior?

“…In residential contexts, people are more conservative and show less impulsivity and shared responsibility [ 129 ]. People mistrust the pre-set schedules of the technologies and find it difficult to understand the broader objectives [ 130 ]. Deeply rooted objections – I cannot live without air conditioning!…”

WHISPER: Wireless Home Identification and Sensing Platform for Energy Reduction

“…The performance of the ARXLR model was compared to several different baselines: a minority vote (MV) algorithm; a non-probabilistic method, based only on the time of day; and a version of the ARXLR model that used ground truth occupancy, instead of predicted occupancy, for the lagged values. In addition to the three accuracy metrics previously described, true and false positive and negative rates were calculated for each, as described by Equations ( 10)- (13).…”

“…Furthermore, many occupants find programmable thermostats difficult to use and fail to take advantage of their features [10]. Smart thermostats, which attempt to learn and respond to an occupant's behaviors and preferences, offer a promising alternative [11,12]; however, usability issues with these devices abound [13], and the energy savings are often much less than promised [14,15]. Thus, accurate occupancy detection, which does not rely on user interaction, has the potential to improve how HVAC systems are operated and to impact building energy use.…”