Adaptive and Intelligent Edge Computing Based Building Energy Management System


The consequences of a more efficient use of energy towards its environmental impact are clear. Some studies (e.g., Guidance on additional energy savings in multi-occupancy buildings) point out that it is possible to reduce electricity use in a building by up to 50% by a number of measures, including some low-cost measures. ‘The EEA Technical Report No 5/2013 Achieving EE through behavior change’ states that combined measures for behavioral change can achieve savings between 5% and 20%, while the report ‘Adding advanced behavioral models in whole building energy simulation: a study on the total energy impact of manual and automated lighting control’ from the National Research Council Canada shows that the inclusion of control and automation patterns can achieve savings up to 40%. Moreover, ecoCASA national project made by UoS has already obtained first saving results in the order 20%-30% with collective challenges and control and automation. Since we will use effective personalized actions for behavioral change in combination with control and automation, it is not unreasonable to assume that the AI-BEMS project will achieve savings in the order or 15% will be achieved in the worst scenario, while savings up to 30% can be achieved in an optimistic scenario. This reduction entails, for example, annual energy savings per household between 2.25 MWh and 4.5 MWh and averaged CO2 emissions between 1 and 2 tons per household, by means of deploying the smart platform in combination with intelligent control and automation.

The economic impact generated by the exploitation of the results is described in the exploitation section, however, the AI-BEMS project directly impact the family/company economies by the savings produced in the energy consumption.

The major expected sociological impact will be the modification of end-users’ beliefs, attitudes and behavior with respect to energy sustainability. This modification is thought in terms of adoption of a new practice for EE and this change will be implemented through a suitable adaptation of persuasion technology (the social machine). To this end, the project will elaborate on the expected change from four perspectives:

  • Comparative Feedback (CF): Using information from control groups to demonstrate the efficiency of the measures proposed is likely to induce group-dynamic changes. This approach clearly outperforms traditional (i.e. single source based) feedback.
  • Ambient Persuasion (AP): The project will also look for the commitment of users providing real time tailored suggestions, that will be a consequence of the scenario where they end-users are immersed. This is aligned with a tailored Ambient Persuasion Model and expected outcomes obtained in terms of social impact will be valuable references for the ongoing metrics developed in the project.
  • Serious Gaming (SG): The competence with other end users and the integration of reputation factors, leverages the interest of participants. These persuasive games are already part of the engagement strategy and will help in increasing the motivation for behavior change .This will enhance long-term actions aimed at teaching energy-saving attitudes.
  • Online Social Network Contagion (OSNC): the dissemination of the results will enhance the spread of a behavioral change over the Internet. Different social network diffusion models will be implemented to boost the spreading of the new paradigm of responsible behavior.