Overview
For my BEng dissertation at DIT, I collaborated with Building Design Partnership (BDP) to develop a custom data analysis tool for analysing Building Management System (BMS) data from a newly opened primary school. The project addressed a real industry challenge: extracting meaningful insights from vast quantities of sensor data.
The Challenge
Chris Croly, Building Services Engineering Director at BDP, articulated the core problem:
“We already have the answers in the form of a large data set but we need to figure out what questions to ask.”
The BMS at Scoil Naomh Eoin (a 16-classroom primary school) was collecting temperature, CO₂, and ventilation data from sensors throughout the building at 15-minute intervals. The raw data existed, but there was no effective way to visualise it, identify patterns, or detect faults in the system.
Technical Approach
I developed the analysis interface in Microsoft Excel using VBA (Visual Basic for Applications). This platform was chosen for its accessibility to building services engineers who may not have programming experience, and its powerful data manipulation capabilities. The development went through three major iterations as I adapted to changing data formats and evolving requirements.
The tool provided four distinct views for analysing the building data:
- Floor Plan View: Spatial visualisation with colour-coded rooms showing real-time conditions
- Graph View: Time-series analysis of temperature and CO₂ levels for individual rooms
- Table View: Heat map matrices showing conditions across all rooms over time
- Data Export: Processed data for further analysis
Floor Plan Visualisation
The floor plan view was the primary interface for understanding building conditions at a glance. Each classroom was colour-coded by temperature (blue for cool, green for comfortable, red for warm) and displayed current CO₂ levels as a percentage of the safe threshold. External conditions including temperature and wind speed were shown alongside the building plan.
This visualisation immediately revealed a critical issue: two classrooms (marked as B1P1Z3 and B1P1Z4) showed constant readings regardless of time or date. The sensors in these rooms had malfunctioned, and this fault had gone undetected until the data was properly visualised.
Time-Series Analysis
The graph view enabled detailed analysis of individual classrooms over selected time periods. Users could plot temperature and CO₂ levels on dual axes to understand how conditions varied throughout the school day and identify correlations between variables.
Analysis of the data revealed that most classrooms maintained CO₂ levels below the safe threshold of 1500 ppm, with occasional exceedances lasting 15 minutes to 2 hours. One notable anomaly was identified: Classroom 1 on 26 October experienced elevated CO₂ levels for 5 hours 45 minutes, suggesting the automated ventilation system may have failed to respond appropriately.
Heat Map Analysis
The table view provided a matrix visualisation showing conditions across all rooms and time periods simultaneously. Cells were colour-coded from green (normal) to red (concerning), allowing rapid identification of patterns and anomalies across the entire building.
Industry Impact
The tool was immediately useful to BDP for reviewing the school's commissioning and identifying faults with the controls installation. Chris Croly confirmed that the interface would be:
- Populated with a full year's data from Scoil Naomh Eoin
- Sent to the Department of Education and Skills to help them understand how schools of this type operate
- Used as a tool to explain to controls specialists how to improve their communication of BMS information
- Expanded to include additional metrics: boiler conditions, water/gas/electrical usage
- Applied to compare data with a second identical school (Scoil Aine Naofa in Ardclough, Co. Kildare)
Industry Testimonial
Chris Croly, Building Services Engineering Director at BDP, wrote:
“Your understanding of computer programming impressed me. This is a rare skill among engineers and I believe it will serve you well in your career... The code was produced in good style with appropriate commenting such that we can further develop and adjust the interface as required.”
Skills Demonstrated
This project demonstrated several capabilities that would prove foundational to my later software engineering career:
- Data Engineering: Handling inconsistent data formats, date parsing issues, and building data pipelines from raw sensor logs
- Software Development: VBA programming with modular architecture, clear commenting, and maintainable code
- UI/UX Design: Iterative interface design responding to user needs and data constraints
- Data Visualisation: Creating multiple view types (spatial, temporal, tabular) tailored to different analysis needs
- Problem Solving: Adapting to changing requirements when data formats changed mid-project
Reflections
This dissertation was one of my first significant programming projects and my introduction to data visualisation challenges. Working with real-world data from a live building system taught me that data is rarely clean, requirements evolve constantly, and effective visualisation can reveal insights that raw numbers obscure.
The collaboration with BDP provided invaluable industry experience. Their feedback on the practical application of my work to real building analysis validated the approach and demonstrated the value of combining engineering domain knowledge with programming skills.
The challenges I faced here, including transforming messy data into meaningful visualisations and iterating on interfaces based on user feedback, would become central themes throughout my career.