Elsevier, International Journal of Human Computer Studies, Volume 137, May 2020
Addressing efficient management of energy has become a central objective due to the scarcity of traditional energy sources and global warming. To cope with this overarching issue, some technological solutions such as Smart Grids, Internet of Things or Demand response are proposed. However, the majority of them overlooks the role of human beings in the equation. Moreover, the very nascent body of research combining human and machine intelligence proposes methods, frameworks, and guidelines which vary depending on the application scenario complicating the selection of gold-standards to ensure seamless cooperation between smart devices and people. Hence, the purpose of this paper is to provide a set of design-hypotheses to devise augmented objects that ally with their users to reduce energy consumption. We expect designers, engineers, makers or even hobbyists in the intersection between technology-enablers (through IoT) and behavioural scientists to benefit from them. To this aim, we describe the results of a long-term study in office-based workplaces, where participants were randomly assigned to different experimental conditions (persuasion, dashboard, and automation) to increase their energy-efficient behaviour. Grounded Theory analysis was applied over qualitative data collected during focus group sessions obtaining five themes around a central category. The resulting themes were linked to design-hypotheses for IoT devices which were then tested through the implementation of a new IoT object also conceived for the workplace.