Following the feedback we received at the Atlanta Science Festival event at Georgia Tech, our team made key improvements to the AQI onboarding prototype in preparation for our second event—this time at Piedmont Park on March 22nd. We focused on refining both the design and content to improve clarity, memorability, and usability, especially emphasizing the health impacts of different AQI levels. We also addressed several minor design bugs that were observed during the first round of testing.
Prototype Improvements
Design Fixes
- Increased the pollutant bubble size to improve tapability on touch screens
- Updated fonts on pollutant cards for better UX consistency
- Renamed the converter tab button to “Tap to Continue”
- Applied a gradient color to “Highest AQI Value” and “Dominant Pollutant” sections
- Made the AQI bubble on the final screen clickable for detailed information
Content Fixes
- Indicated that the data shown is historical sample data from a specific date
- Added pollutant unit information to increase clarity
- Included face illustrations to help users remember health impact categories
- Added an AQI scale and visual indicator to explain what a value of 136 means
- Included instructions on how to check daily AQI information
- Clarified that the AQI displayed is not real-time
- Added details about sensitive groups and who they include
Event Highlights
Around 100 visitors engaged with our setup, which featured the newly updated AQI onboarding prototype displayed on an iPad and a large TV monitor showcasing Atlanta’s 10-year AQI trends. Due to the side-by-side arrangement, many attendees first gravitated toward the data visualization screen before interacting with the prototype—leading to some interesting cross-screen comparisons and questions.
We asked participants what they learned, whether they had further questions, and what they would like to explore next. Their feedback revealed both curiosity and critical thinking about air quality and its impact on daily life:
Key Audience Insights
- Wanted to compare AQI across birthdays, seasons, and significant dates
- Asked to include weather data (e.g., wind, temperature) to contextualize AQI changes
- Surprised by the overall air quality in Atlanta and curious about health correlations
- Requested a short video to explain how pollutants are measured and converted
- Asked whether pollutants beyond the six presented are also monitored
- Suggested comparing AQI trends across U.S. cities and internationally
- Expressed interest in predictive AQI modeling, similar to weather forecasting
- Asked about the connection between AQI and hospital visits or respiratory conditions
- Suggested linking extreme AQI days to causes (e.g., wildfires, traffic) and policy changes
- Wondered how AQI changes from day to night and between pollutants like PM2.5 and ozone
- Recommended showing average counts of green/yellow/red days through line charts
- Proposed connecting each pollutant level with suggested daily actions
What We Learned
Visitors were especially intrigued by the idea of forecasting AQI using AI, and they expressed a strong desire for contextual data—like what might have caused a spike on a certain day, and what actions they could take in response. There was also a clear interest in not only understanding AQI trends in Atlanta but comparing them across cities and seasons.
These insights will directly inform our next design iteration of the AQI trend screen, which is shown after the onboarding prototype. We’re exploring how to present more actionable information and predictive data while continuing to keep the experience engaging and accessible to K–12 audiences.
Reflection
Across both ASF events, our team learned how deeply audiences care about air quality when given the tools and context to understand it. We saw how interactive design, when paired with meaningful visual storytelling, could spark conversation and raise awareness—especially among younger participants.
Most importantly, we recognized that people want to make informed decisions about their everyday activities based on AQI, and they’re eager to know how to access, interpret, and act on this data. As we prepare for the next round of iteration, these insights will be essential in shaping a tool that not only informs but empowers.