Visualizing the user in the building data

Julien Nembrini
Denis Lalanne
Roberto Sanchez
Project status: 

With the advent of Big Data, the means of collecting data have been improved in the last decade. However, the extraction of information out of it becomes more and more difficult; the danger of getting lost in data increases.

The combination of Human Computer Interaction and information visualization has the potential to let humans explore data sets interactively to discover unforeseen patterns, for instance detecting failures in systems or understanding specific optimization potential.

The goal of this master project is to develop an interactive tool to allow expert users to explore multivariate building data. Current performance monitoring systems for buildings use for their dashboards classic data visualizations such as time series and bar charts.

Building on the outcomes of the project "Visualization of multivariate building data", and using real data coming from industry and the SmartLivingLab, which include variables such as power consumption, room temperature, humidity, CO2, etc, the project will combine machine learning and statistical methods with visualization to:

1. address the problem of quality and quantity of building data coming from existing sensing.
2. analyse the influence of user behaviour on the building performance (Human Building Interaction).  

Following the focus on HCI methods, emphasis will be put on the implementation of a prototype and its evaluation through user studies

Miller and Schluter (2015) Forensically    discovering    simulation    feedback knowledge    from    a    campus    energy    information system