Aura
Translating environmental data into actionable productivity metrics
Data visualisation · Web design
Overview
The University of Sydney’s IEQ Lab gave us a year of real office sensor data. The brief was to find who would care about it, then build an interactive web app that turns it into a story they can act on.
I built Aura, a product page that combines a narrative landing page with a working dashboard, translating raw sensor readings into productivity and health insights for non-technical users. I ran the full process end to end, from research and conceptual development through design and front-end development.
Project type
Academic
Timeline
Feb - May 2025
Tools
Figma
Miro
HTML, CSS, JavaScript
D3.js
Excel
ChatGPT & Claude
Achievement
Context
What is Indoor Environmental Quality (IEQ), Sick Building Syndrome (SBS) and SAMBA?
Indoor Environmental Quality (IEQ) refers to the conditions inside a building that affect how people feel and perform: air quality, temperature, lighting, and noise.
When these factors fall outside healthy ranges, the effects show up as symptoms like headaches, fatigue, and difficulty concentrating, collectively known as Sick Building Syndrome (SBS).
SAMBA is a low-cost distributed sensor system developed by the University of Sydney’s IEQ Lab. It tracks 10 environmental variables across multiple zones and floors at 5-minute intervals, generating a detailed picture of how a building’s conditions shift throughout the day and across seasons. Below is a snapshot of the raw data I had to work with throughout the project.
the Problem
The data exists. The people who need it can't use it.
The brief covered IEQ broadly, so I first needed to find the story inside it. How does IEQ affect the people inside a building? The research kept pointing to two groups that were blocked by the same problem.
Office workers
IEQ has a physical impact on their health and performance.
Low IEQ
SBS
IEQ exposure has been linked to Sick Building Syndrome symptoms, difficulty concentrating, and memory lapses.
Employers
The impact is financial when employees’ productivity increases. However, the numbers are only clear if someone can connect building conditions to performance.
potential annual savings from 5% productivity uplift.
How might we…
make IEQ data accessible and meaningful for office workers and employers so they can understand its impact on productivity and health?
Exploring the data
Breaking down data step-by-step
To test whether visualising IEQ impacts was feasible, I started with the SAMBA dataset and worked through a four-step analysis:
1
Selected four variables with the strongest research-backed links to productivity and health: temperature, CO2, sound (dB), and light (lux).
2
Simulated productivity gains and SBS risk scores using thresholds from published research (Felgueiras et al., 2023; Clements-Croome, 2017; Lowther et al., 2021; Boyce, 2010; Lee et al., 2016).
3
Built the simulation in Excel using IF formulas to score each reading against its threshold, then normalised the results.
4
Used pivot tables to analyse the data top-down: all zones first, then specific zones and time-of-day patterns.
Insights
What the data revealed
From the 4-steps top-down analysis, I derived three insights that shaped the concept.
Performance varied sharply by zone
E.g. Productivity gains ranged from 2.86% in Zone 243 to 11.47% in Zone 403; and zone. A building-wide average would have hidden that difference.
Different zones had different problems
Lighting was the main SBS risk driver across most of Floor 0, except in Zone 402, which showed almost no lighting risk. The causes were not consistent across the building.
Conditions changed throughout the day
In Zone 243, lighting and noise contributed to SBS risk consistently, but CO2 spiked during peak hours (9 AM to 5 PM) and aligned with dips in productivity. Even zones with acceptable overall CO2 levels still had time-based risks.
Key takeaway
IEQ risks are real and measurable, but they change by zone, factor, and time of day. Static building averages hide the issues that matter most. The visualisation needed to show zone-specific, time-sensitive, factor-level detail.
Aura: An IEQ analytics dashboard for productivity and health-conscious workplaces
The deliverable was a product page with a working demo and the narrative around it, so I split the page into two halves: the first half builds the case for why IEQ matters, while the latter lets users explore the data themselves.
Connecting symptoms to IEQ
Most people don’t know what SBS is or how IEQ relates to their symptoms. I opened the page by naming that gap, then used an interactive D3.js network chart to make those relationships visible and explorable instead of explaining them in static text.
Pinpoint the cost of inaction
Awareness alone rarely changes behaviour. I added two visualisations to make the business case clearer: potential US savings from reduced absenteeism, and cognitive gains of up to 101% under better ventilation.
Raw vs translated
I placed raw SAMBA readings next to Aura’s translated output to show the value of interpretation directly. That comparison makes the product’s role clear in one glance.
Floor overview
I wanted the dashboard to answer the biggest question first, so it opens with a floorplan heatmap and a floor-level summary of health, productivity, and absenteeism. This gives users an immediate overview before asking them to inspect individual zones.
Zone drill-down
From there, users can click into a zone to see four key metrics, hourly IEQ trends by variable, and a pie chart showing which factors are driving SBS risk. I structured this to mirror the analysis process I used in the data exploration: start broad, find the problem area, then isolate the cause.
RETROSPECTIVE
Some learnings along the way
Getting comfortable with messy data
Before this project, I was intimidated by large datasets and spreadsheet-heavy analysis. Working through a year of sensor readings across multiple zones forced me to get comfortable with pivot tables, IF formulas, and data modelling. That confidence now carries into how I approach data-informed design decisions.
Using AI tools to bridge knowledge gaps
I used ChatGPT and Claude to help structure the data simulation approach and write functional code for the landing page and dashboard. It didn’t replace the need to understand what I was building, but it helped me move faster through unfamiliar territory in coding and data analysis.
