E-SCOUT

The Environmentally Sustainable Computing User Trial of the Green Algorithms Initiative

Tools such as the Green Algorithms online calculator have proven useful for researchers, research software engineers (RSEs) and organisations to understand the environmental impacts of their work, but some important questions remain: To what extent does carbon tracking incentivise green computing practices? Or how can we make sure it does?

Study aims

To empirically evaluate the effectiveness of carbon reporting tools in reducing the environmental impacts of scientific computing, we are organising a study called E-SCOUT. Set up as a multi-centre, pragmatic, cluster randomised controlled trial, E-SCOUT is designed to evaluate the effectiveness of a carbon monitoring dashboard for high-performance computing (HPC) on participants’ behaviour, pro-environmental attitudes, and awareness of the environmental impacts of scientific computing. Responding to the limitations of earlier short-term eco-feedback studies, it will run for six months (plus an initial pilot phase) and involve at least six research-performing organisations.

The carbon reporting tool that will be used by E-SCOUT participants is the newly developed Green Algorithms dashboard, which estimates and visualises the energy consumption and greenhouse gas emission statistics of individual researchers, research groups, and departments within participating organisations.

The research protocol

E-SCOUT’s research protocol is work in progress and will be refined based on the outcomes of the pilot phase. Generally, E-SCOUT will be designed as a mixed methods study, combining quantitative analysis of HPC data with qualitative methods, such as semi-structured interviews. Within participating organisations, we are envisioning the division of participants (on a departmental level) into three intervention groups: (1) dashboard only, (2) dashboard and community support, and (3) control (no intervention). Community support does thereby refer to a package that includes additional resources, access to the ESCS community platform for Environmentally Sustainable Computational Science, drop-in sessions, and/or departmental ‘champions’.

We are also setting up an advisory group that brings together people with diverse perspectives for guidance and decision-making.

Data and privacy

The Green Algorithms dashboard uses a layered approach to privacy, with individuals only having access to their own footprint data and those of the rest of their research group, and aggregated metrics outside of their group. The data collected by the dashboard will remain with the participating organisations; individual-level HPC usage data will only be shared with the E-SCOUT research team in anonymised form. For the collection of qualitative data, informed consent will be obtained from all participants.

How can I get involved?

If you work for a research-performing organisation — such as a university — that uses HPC, and would be interested in getting involved, please register your interest here. It would be great to hear from you, and we will reach out with further information once the study has been reviewed by our Ethics Committee.

If you have any questions about the study, please do not hesitate to contact Dr Christina Bremer ([email protected]).