The Value of Data

One of the first steps in establishing a living lab for sustainable energy use is to gather and examine existing data relating to energy consumption in the building where the living lab is situated. There are a few good reasons for doing this. Firstly, existing data can provide baseline measures of energy usage, which could be useful at a later stage when trying to assess the impact of any interventions. By understanding what data are already available, it is also possible to assess what additional data may need to be collected over the course of the living lab. Additionally, visualising and presenting relevant data to building users can be a great way of stimulating discussion and exploring issues surrounding energy use.

iLights at the Assembly Rooms

At the Assembly Rooms, a historic cultural venue where we are running a living lab with the City of Edinburgh Council, we were excited to find that the building’s lighting is programmed and controlled using a network-based iLight system. Consequently, all lighting use, down to the level of individual switches, can be recorded and monitored – a fact made all the more interesting by the spectacular lighting set up in the Assembly Rooms, which consists of ornate chandeliers made up of hundreds of dimmable bulbs.

The production staff at the Assembly Rooms have direct access to the iLight system through a software interface, enabling them to program individual lighting controls and scenes. However, the software provides little functionality for monitoring and visualising lighting use. The the only way of logging lighting use is to save the network messages, which are somewhat difficult to interpret (see the figure below). Consequently, we decided to develop some more user-friendly ways or exploring and visualising the lighting data.


iLight log data

Dynamic Floor Plan Visualisation

The first visualisation, shown in the video below, uses a floor plan of the Assembly Rooms with appropriately sized circles overlaid on top to represent individual lights. After selecting a specific date, the visualisation tool allows you to view the lighting use over the course of that day by clicking play. Since most of the lights in the Assembly Rooms are dimmable, the brightness (or ‘yellowness’) of each circle denotes the current light level of that particular light. Whilst the visualisation is playing it is possible to hover over individual lights to see their name and brightness at that particular point in time.

Static Multi-day Visualisation

The previous visualisation provides a good way of viewing the spatial arrangement of  lights and their use over the course of single days. However, it is not so well suited for viewing patterns in use over longer periods of time. Therefore, a second visualisation tool was developed in order to plot lighting data for specific lights and areas over user-defined date ranges. The tool is demonstrated in the video below. The visualisation consists of a grid, where the rows represent individual lights, and the columns are points in time. The colour of each grid segment indicates the light level of the corresponding light at a specific point in time.

These tools are early prototypes to show what kinds of visualisations might be possible with the lighting data we have acquired. As our living lab progresses, similar visualisation tools could be co-developed with building users to assist them in monitoring, understanding, and perhaps even changing their behaviours relating to lighting use. For the latter to occur, it may also be necessary to know something about why the lights were being used at certain points in the day. Consequently, one of the next steps is to gather occupancy and events schedule data to incorporate into a lighting use visualisation.

The visualisations described above were created using Python and the Bokeh interactive visualisation library (technical details to be described in a forthcoming post). The code and files needed to run the visualisations are here.



Visualising lighting data at the Assembly Rooms
Tagged on: