From Data to Decision : Software-Driven Energy Awareness and Eco-Feedback Mechanisms in the Public Sector

Sammanfattning

The digitalization of the public sector has led to an extensive growth in IT infrastructure, significantly increasing the energy footprint of organizations like Region Blekinge. While hardware is usually capable of reporting energy metrics, this data often remains “invisible”, heterogeneous, and inaccessible to strategic decision-makers. Existing research suggests that “eco-feedback”, making consumption visible, can drive behavioral change, yet the public sector lacks the necessary automated tools to translate raw telemetry into actionable, visualized insights. This thesis seeks to bridge the gap between heterogeneous, technical data, and organizational energy awareness. The primary objectives were to: (1) investigate motivations and barriers for energy monitoring in a public organization, (2) develop an automated software artifact capable of extracting and standardizing metrics from heterogeneous network hardware, and (3) evaluate the extent to which data visualization motivates organizational change. The methods used to bridge the research gap consisted of interviews with technical and political stakeholders, as well as technical personnel in Region Blekinge, in order to investigate motivations and barriers for energy monitoring at a larger scale. The monitoring artifact was developed with an agile method consisting of a product owner and user stories, and was constructed using both components built from scratch, as well as already existing open-source software ready for deployment. The artifact and its general usefulness, as well as how the proposed software could be used to implement eco-feedback mechanisms and promote energy awareness in Region Blekinge, was later further evaluated through a quantitative survey, with the employees of Region Blekinge’s IT department as participants. Despite the challenges with lacking protocols and heterogeneous hardware, the artifact successfully standardized power consumption metrics and displayed them in interactive Grafana dashboards. The empirical study revealed that 83.8% of participants found the visualizations intuitive, and 63.3% believed the tool could significantly increase the organization’s energy awareness. This thesis demonstrates that there is a severe lack of automated tooling for measuring energy consumption in the public sector, but also a considerable demand for such a tool to exist. By providing a single source of truth for infrastructure telemetry, the developed artifact empowers stakeholders to make data-driven decisions regarding procurement of hardware and operational efficiency. The findings suggest that visibility is a large contributor to promoting the “energy-saving mindset” required to meet modern environmental goals.