Scope
Souza et al. argue that monetary incentives are less effective in the long term and do not provoke a stable change in users’ habits and preferences. It is considered that user behaviour can be modified towards carbon footprint reduction through informative systems that do not heavily change the user consumption pattern.
Summary
A user-centered decarbonisation strategy is envisioned that includes energy system optimisation schemes with an AI-augmented user behavioural intervention mechanism. User engagement is achieved through the assistance of users in making carbon-efficient choices while considering their habits and preferences, thus mitigating the utility loss associated with the participation. An interactive visualisation is also employed that identifies the challenges and opportunities in formulating carbon-efficient policies targeting various settings and user groups.
Insights on the problems discussed can be extracted from this work. For instance, it is explained that renewable sources of generation are intermittent and as a result the carbon emissions are associated with the mix of generation sources at any given moment. In fact, the carbon intensity (grams of CO2 equivalent per watt) varies notably through time and exhibits significant discrepancies across regions. As a result, the task of designing energy-efficient activities involves the consideration of the local energy system. The recent emergence of carbon information services that provide live data on the carbon intensity of generation, may empower users to make informed choices on their consumption patterns. It is elaborated that monetary incentives have proved to be instantly effective in reducing consumption across the user base, however, maintaining this behavioural change in a permanent manner personal habits and routines should be modified employing the behavioural science. Behavioural change should be effectively promoted while minimising the respective user burden. To this end, an AI-assisted decarbonisation framework is proposed that is composed of three main components: 1) energy system optimisation modelling, 2) AI-augmented user behavioural intervention, 3) analytics and visualisation.
The first component takes into consideration the usage profiles of the appliances and the grid status and then utilises data engineering approaches to provide a sum of optimal and near-optimal energy profile settings for the household appliances. The second component is fed with data collected from different representative surveys and public reports. Post-processing of this data can enlighten behavioural aspects of the users and provide an estimation of their motivation to reduce their carbon footprint. Clustering algorithms are also proposed to discover profile patterns among the users. The third component concerns a visualisation scheme that communicates to the user the effects of his carbon footprint. Continuous adaptation is proposed by user feedback to create recommendations on the provision of suggestions. The framework essentially suggests that by employing aggregates of users’ behavioural preferences and routines, and energy usage for daily activities, enabling the visualisation of carbon data, demographics, and energy consumption habits, the proposed dashboard can provide actionable insights to trigger societal awareness and behavioural change.
Relevance for EXIGENCE
Souza et al.‘s publication discusses what incentivises behavioural change among consumers and describes the challenges in using automation technologies with recommendation systems to sustain consumer engagement in decarbonisation efforts. Insights can be derived for EXIGENCE on the key considerations regarding user approach for sustainability purposes.
Souza, M. Shenoy, and C. Zakaria, “Empowering User-Centered Carbon Management: Bridging Individual Preferences and Sociotechnical Advancements,” in Proceedings of the 10th ACM International Conference on Systems for Energy-Efficient Buildings, Cities, and Transportation (BuildSys ’23), Istanbul, Turkey, Nov. 15-16, 2023, pp. 1-7.
