Background
The IDLES (Integrated Development of Low-carbon Energy Systems) programme (2018-2023), funded by EPSRC and hosted by Energy Futures Lab, develops soft-linked modelling tools to support long-term energy system planning for net zero. Within this, the Gridlington project integrates several tools to showcase research, focusing on the role of demand-side response in grid operations at both system and local levels by linking four subsystems (models of investment in the electricity system, electricity networks, agent-based demand models, and intra-day balancing markets). The current challenge is building a visualisation framework to connect subsystem outputs and make data more accessible. Once developed, this will allow researchers to present scenarios interactively to visitors and stakeholders, providing a key output of the IDLES programme and helping maximise the real-world impact of its research.
Our Contribution
Visualising the Gridlington scenarios brought a number of unique challenges. The Gridlington Demonstrator room contains 8 screens connected to 4 different computers, the goal was to be able to control what appears on those screens from one central point for displaying scenarios to stakeholders. This was achieved using the Imperial-developed Open Visualisation Environment tool - a software designed for running large-scale visualisations in the Data Observatory in Imperial's Data Science Institute. Our use case for this tool was unique and required collaborating with the OVE development team to ensure it would work for us. The gridlington-vis tool included web-based dashboards displaying data on all 8 of the screens as well as a controller page that would be connected to via a tablet device used by the demonstrator.
There was a second component to this project that was to create a central point through which the visualisation tool could communicate with all of the models. This gridlington-datahub was developed using FastAPI and deployed in a docker container so that the models and visualisation tool could communicate with it via RESTful APIs.