DSM based on collective intelligence (CI)

Purpose and goal:

The project was designed to enhance the design of a novel demand side management (DSM) system based on collective intelligence (CI), called CI-DSM, in a way to better account for complex interactions in urban areas, focusing on user comfort/health, especially for elderlies. By integrating a decision making approach based on reinforcement learning (RL), this project could enhance the design of CI-DSM to increase user comfort and flexibility energy system, resulting in improved sustainability and resilience of urban energy systems.

Expected results and effects:

Integration of CI and RL in the energy management has resulted in a collaborative system that quickly responds to the variations of the outdoor environment without the need for heavy/expensive calculation and ICT solutions while minimizing the need for data transfer. For example, in the COLLECTiEF project, we are working on implementing such solution on a Raspberry Pi.

Such a system can become an affordable solution towards increased sustainability of energy solutions and climate change adaptation of urban environment. Approach and implementation:

The developed collaboration and results will be further expanded through the ongoing and recently funded projects as well as the research applications. We are running relevant European and national project together with NODA. Moreover, closer collaborations are planned with Hiroshima University (and in general Japan) through applying for affiliated positions.

We ran a joint workshop in MIRAI 2.0 event in November 2022, presenting some of the results of our joint collaboration. Currently, we are working on joint publications and planning for further collaboration and exchange.