With smart IT systems, modern buildings can manage heating, cooling and lighting on their own – on a sunny day they will, for example, automatically close the blinds. But there are many factors influencing a building, and getting the programming completely right is not an easy task. Construction specialist Paul Wurth Geprolux has joined forces with the University of Luxembourg to use big data analysis in order to get an in-depth understanding of how buildings behave and thus improve its automation systems.
With its extensive experience of managing major building projects and expertise in the technical engineering of buildings, Paul Wurth Geprolux knows the potential for enhancing existing systems. “In every up-to-date building, large amounts of data are permanently generated by thousands of interconnected devices that operate various systems: heating, air conditioning and so on,” explains Paul Schummer, Project Engineer at Geprolux. “This data could tell us a lot about what is going on in the building, how the automated systems respond and what the results are – for example, whether the building is capable of keeping a constant, pleasant temperature without excessive use of the heating or air conditioning.”
Analysing these enormous series of data is, however, a daunting task that requires advanced skills in big data analysis, algorithm development and innovative data treatment techniques. In 2015, Paul Wurth entered a four-year research collaboration with the SnT, the Interdisciplinary Centre for Security, Reliability and Trust of the University of Luxembourg. The two partners have started by collecting building data in a large school in Luxembourg for almost a year, and more than 18 million data sets are currently being analysed by the SnT. “Applying our models on a real-life case and working with Paul Wurth’s smart building experts allows us to optimise our algorithms in a way we could not have done on our own,” says Jacques Klein, Senior Research Scientist at the SnT.
And the knowledge is already being put to practical use. “We are developing visual dashboards that provide an easy overview of what goes on in the building and help detect anomalies,” says Mr 16 Schummer. “Our goal is to help facility managers identify situations when they need to intervene, by reprogramming certain automatic actions for instance. We also want to help everyday users understand why certain actions have been programmed so they don’t manually override them in a way that has a negative impact on the use of resources. If the blinds close when the sun is not yet very strong, it might be to avoid that the building quickly gets too hot and uses more energy for cooling down than what would otherwise be needed.”
The strength of the collaboration lies in bringing together the complementary skills of the partners. “We are confident that the combination of our two worlds will lead us to fruitful results,” Mr Schummer concludes.