Data Science Use Case: Niko Home Control
Updated: Nov 8, 2021
What is Data Science?
Data Science focuses on the collection, modelling and interpreting of company data. This helps organizations reach better conclusions and make better decisions as a result of this data.
Earlier this year, Niko, a leading supplier of home automation, carried out a pilot project in a limited number of student houses. The aim: to see how Niko Home Control can benefit the landlords of these properties as well as the students who live there. First and foremost, all the data collected by the sensors and meters needed to be saved in a structured format to enable further processing. Next to that, this now structured data was analyzed further and then visualized so that existing hypotheses could be tested and new insights could be generated.
CANGURU supported Niko throughout this whole process. An automated data pipeline was set up to save the unstructured data, originating from various sources, in a central, cloud-hosted SQL database. The generated datasets were therefore directly available for further processing. This data was thoroughly explored to get clarity on which information was available and – equally as important – which data was still missing.
After analysis, the great amount of incoming data was reduced by half. Not only is this important so that the useful data can be separated from the unnecessary data, but this also allows the scalability and sustainability of the monitoring in the longer term.
Finally, many different perspectives were delivered in order to visualize this data, both statistically as well as dynamically.
CANGURU is a young and innovative consulting company based in Mechelen. The data science team helps companies reach better conclusions and make better decisions as a result of “smart data”. We implement the right tools to gather and compile qualitative data. We subsequently bring this information into the light which should help the company take the correct decisions.
Want to know what we can do with your data? Let's talk!