People's interaction with future autonomous energy systems in their everyday lives at home

Abstract

Intelligent agents that sense and respond our continuing daily activities autonomously are becoming increasingly ubiquitous, and consequently transforming our lives. Domestic energy use accounts for a significant portion of national energy consumption in many countries, and is an important domain that intelligent agents may provide great benefit for us, by enabling a much more efficient energy utilisation. Whilst there are many algorithms developed for autonomous agents to assist people in managing their energy consumption at home, to date, there have been very few studies that examine human interaction with autonomy in the wild. Hence, there is a significant gap in our understanding of how people would react to and interact with autonomous agents in their everyday lives. This thesis aims to close this gap and help us to better understand how to design user interaction with future autonomous energy systems. To this end, we focus on people's perceptions of and interactions with two agent-based energy management systems that we designed and deployed based on envisioning future energy scenarios, and evaluated these systems through field studies. We represent implications for the design of future intelligent energy systems based on the results of our field studies.

The first system focuses on energy tariff switching. The decision of which energy tariff (i.e., energy pricing schema) to select is a challenging task for today's most households. Energy companies offer many different tariffs (e.g., standard, time of use and realtime tariffs) and it can be difficult to know which will be the most tailored to your consumption profile. Furthermore, the changes in the households' consumption and tariff rates increase the likelihood of ending up with a wrong tariff decision. To this end, we first focus on a future scenario where autonomous agents embedded in households have the ability to switch the energy providers daily, based on their offered rates and the households' consumption routines. To instantiate this envisioned scenario, we designed and developed two prototypes of a novel home energy management system called Tariff Agent, which monitors household energy consumption, as well as available energy tariffs, and therefore calculates the best tariff, and (optionally) automatically switches to it. Both Tariff Agent prototypes offer flexible autonomy by which users can shift the system's level of autonomy in switching tariffs among three options: suggestion-only, semiautonomous and fully autonomous, whenever they like. The first prototype was used by 10 UK households for 14 days. The findings from both quantitative and qualitative results of this first field study show that at least some people are ready to embrace software agents to manage their energy tariffs on their behalf as long as the agents reduce the hassle of tariff switching and maintain their budget. The results also indicate that although the users showed trust in Tariff Agent to control their tariff, they were still keen to monitor its performance. The second prototype was built based on the results of the first study and, differently from the first prototype, users are enabled to change the frequency of system reports that were previously sent once on each day of the study. To examine user interaction with the system for longer terms, the second study lasted 42 days and involved 12 UK households. The findings based on a thematic analysis show that flexible autonomy is a promising way to sustain users' engagement with smart systems, despite their occasional mistakes. The findings also suggest that users take responsibility of undesired outcomes of automated actions when delegation of autonomy can be adjusted flexibly.

The second system focuses on home heating. Home heating is a primary portion of energy expenses and therefore it is an important issue for residents. A number of smart thermostats have been introduced to customers to automate heating control on their behalf with the purpose of increasing the home's energy efficiency. However, none of these thermostats take into account energy prices that may vary based on residents' energy tariff. Hence, the second future energy scenario that we focus on envisions a smart thermostat that automates home heating control when energy price varies in realtime. To do so, we implemented three different smart thermostats that automate heating based on users' heating preferences and the real-time price variations. We evaluated our designs through a field study, where 30 UK households used our thermostats to heat their homes over a month. Our findings through thematic analysis show that the participants formed different understandings and expectations of our smart thermostat, and used its different features in various ways to effectively respond to real-time prices while maintaining their thermal comfort. Based on the findings, we present a number of design and research implications, specifically for designing future smart thermostats that will assist us in controlling home heating with real-time pricing, and for future intelligent autonomous energy systems.

More information

Identifiers

ORCID for Sarvapali Ramchurn: orcid.org/0000-0001-9686-4302

Catalogue record

Export record