Fontys IT Dataproject

Project description

R10 is a new building with sophisticated sensors. These sensors have already been used for measuring and saving data in 2020, but nothing has been done with this data yet. Fontys, IT wants to use this data for finding patterns that help to improve the education.

The main question for the first part of this project was:

“How can R10 data and external data be used to obtain information for Fontys IT?”

Our assignment was to design a database and dashboard with external and internal data of (a part of) the Fontys R10 building, to enable Fontys IT employees in Eindhoven to draw certain conclusions regarding the number of present students in R10. Next groups can use this database and dashboard, while working on the following question:

“How can Fontys IT use data about buildings (sensors), occupancy and external data to find patterns that help to improve the education?”

Context

With new technologies, buildings can become more sophisticated. One of those buildings is R10 from Fontys University of Applied Sciences in Eindhoven, which has extra features like sensors to measure light, CO2-emissions, and Wi-Fi utilization. The project group Novo Data will combine these datasets with external data, to make a database with cleaned information.

These sensors have already been used for measuring and saving data in 2020, but nothing has been done with this data yet. Fontys IT wants to use this data for finding patterns that help to improve the education. Both internal and external data are needed to find correlations and influences, so the goal of Fontys IT cannot be achieved by only using the data of the sensors. Moreover, the collected data needs to be transformed to information for finding the correlations and influences.

Results

Azure Calculations

Fontys IT asked us to use Azure for the project. The platform itself is free of charge, but some components can have a monthly fee(pay in advance or pay per use). To predict the costs for the usage Azure we made a cost analysis. The analyses are executed on TR Level 2, we did applied research by using excel to calculate the cost with the help of the Azure calculator.
The results are validated by discussing the findings, in a form of presentation, with our stakeholders and our client. They told us to also take in account that more data can be processed, and maybe new components will be used in feature projects.

Research Document

This document contains our technical designs for the Azure architecture, research on optimal values for the sensor data, what data we are going to store and designs for the dashboard. Validating of these results is mainly done by peer-review with the stakeholders. This way we could improve our designs and validate our findings for the data. We also used other research to confirm our findings for the optimal light, CO2, and temperature levels.
The added value for the research document mainly resides on fundamental research and applied research to find feasible designs and data. Maybe it can also be added value for a proof of concept because the research document also contains some choices, we made for the dashboard prototype.

Azure databases and function

Our Azure components, or resources in terms of Azure, are the lifeline of our products. We make use of the Azure function to implement and test our code(TRL level 4 and 5) to get data from different sources and the database to store the data. By validating and testing the pipeline to transfer the data we make sure that the correct data is stored.

The configuration of the pipeline and database was mostly validated by the stakeholders. The client needs to validate the value of the data that is collect through the sources. The data needs to be a value for next projects. Minor testing and validation were also done by the Fontys IT project group, mainly through git or docker.

Analytics (Dashboarding)

Figure 1 Linear graph to test relation between CO2 and personcount
Our dashboards are developed using Microsoft PowerBI. By applying our research on possible relations between the data (TR Level 1 t/ m3) and connecting the database form Azure to our dashboard, we found some connections between the data. By creating these dashboards, we could show the client there is for example a connection between the number of persons in the data and the CO2 levels.