A study on how a dashboard for energy data visualisations can be designed to be usable and inspire pro-environmental behaviour in an industry setting

Abstract: An essential part of combating climate change is to reduce energy consumption. As the industrial sector was accountable for approximately a third of the total global carbon dioxide equivalents in 2019, it is crucial for industries to manage their energy usage. Research within the field of eco-feedback shows that visualising energy data can inspire energy reductions and promote knowledge. The thesis investigates how a dashboard for energy data visualisations for the industrial sector can be designed to be usable and to inspire behavioural change for its end users. Usability is researched in terms of initial learnability and subjective satisfaction. Perceived behavioural change is studied through changes in users' awareness and attitude towards energy management.  A prototype of a dashboard for energy data visualisations is developed through three iterations of the HCD process. During the first iteration, a paper prototype is developed, and evaluated through a workshop. During the second iteration, the paper prototype is translated into a digital prototype, which is then translated into a software prototype. The software prototype is evaluated through end-user tests at the end of the second iteration. During the third iteration, the software prototype is improved, based on the input from the previous evaluation, and then evaluated through end-user tests. The usability of the software prototype is evaluated through the think-aloud method and a SUS questionnaire. Perceived behavioural change is evaluated by interviewing the test users about their change in attitude and awareness after using the dashboard.  The thesis finds that the dashboard prototype was easy to use and interpret. The following design choices were concluded to contribute to the usability in terms of initial learnability and subjective satisfaction: visually separating information, information adjacency, adding explanatory texts, using colour as a visual encoding, using a colour scheme that provides contrast and consistency, filtering data through multi-select, providing different options for visualisation, using interactivity for further data exploration, giving the user response to actions, considering chart junk and the data-ink ratio, and using unambiguous icons. In addition, the thesis concludes that the dashboard inspired changes in both the user's awareness and attitude towards energy management. The design choices that were proven to affect the user's awareness and attitude were: letting the user choose their preferred unit to represent the data, presenting the user with tips, providing internal and external data comparisons, visualising multiple energy-related key figures, allowing for historic comparison, reminding the user of upcoming activities, presenting a list of scheduled measures and providing feedback of a prognosticated proposed measure in terms of its intended effects.