Wabash Lights Data Viz
Project Goal: Design a data visualization for The Wabash Lights, an interactive public arts platform in downtown Chicago.
Role: Data designer
Skills: Adobe Illustrator, Adobe Photoshop, Chromatik
Initial Observations
Engagement is passive and brief:
- Most people are not actively stopping to watch the lights; instead, glancing at them while walking by.
- The current installation functions as a background visual element, rather than a focal point, indicating a need for swift, eye-catching moments rather than designing for long engagement.
Viewing angles are limited and crucial:
- The lights are best observed from each side of the street on the sidewalks.
- The design must prioritize strong, intentional side-view composition that respects and enhances the experience from both sides equally.
Timing and animation impact perception:
- The lights vary between static and animated displays, decreasing engagement at moments of fixed material.
- Immediate animation will help bring visual variation to the lights and capture attention.
Concept Sketches
Utilizing the data compiled in the 311 Service Requests Rodent Baiting dataset by the cityofchicago.org.
Sketch 1: Simple Bar Graph
- The simple bar graph allows for clear representation of data, but lacks uniqueness and creative visualization, even within the contraints of the Wabash Lights.
Sketch 2: Stacked Bar Graph
- Feedback suggested considering how low data variance might weaken the visual impact of this design.
Sketch 3: DNA-Style Viz.
- This abstract design was well-received conceptually, but feedback highlighted concerns about visual clarity and the limited time range on the y-axis to represent individual cases.
Sketch 4: Line Graph
- This sketch was seen as well-suited to the medium's constraints, with feedback focusing on refining the system assessing data severity.
- This sketch uses color-coded line graph elements for clarity from any viewpoint.
Sketch 5: Cluster Plot
- This sketch was well-received, with the cluster plot concept praised for its visual interest and potential effectiveness across the light strips.
- This sketch expands on Sketch 4, with a cluster plot that reveals seasonal rat trends across multiple years.
Prototype 1
Utilizing the 2015-2018 rodent baiting requests data compiled in the 311 Service Requests Rodent Baiting dataset by the cityofchicago.org.
Line Graph:
- While the color choices may need to be brightened for better visibility on the LED strips, the data is clearly presented and the color-clarification concept is strong.
Cluster Plot:
- This well-received prototype was praised for its clarity, vivid colors, effective clustering to show rat activity levels, and the appealing texture it creates.
Prototype 2
A Hybrid Line Graph and Cluster Visualization for Side-By-Side Comparison
Utilizing the 2011-2014 rodent baiting requests data compiled in the 311 Service Requests Rodent Baiting dataset by the cityofchicago.org.
- Feedback highlighted effective color use and promising side-by-side concept.
- Adjustmnets needed for animation timing, speeding up the line graph and allowing more time for the static cluster plot, for clearer visual connection and improved cohesiveness.
Prototype 3
Utilizing the 2011-2018 rodent baiting requests data compiled in the 311 Service Requests Rodent Baiting dataset by the cityofchicago.org.- The third prototype showed strong improvements in animation speed, leading to better visual clarity and cohesion between the side-by-side line graph and cluster plot.
- During an in-person prototype test with the lights, some colors, particularly blue and yellow, appeared too faded. These colors will need increased saturation for improved visibility on the actual installation.
Final Data Visualization
Utilizing the 2011-2018 rodent baiting requests data compiled in the 311 Service Requests Rodent Baiting dataset by the cityofchicago.org.
- Line graph animation (2011-2014): Each year of rat baiting request data is shown across the full span of the Wabash Lights, animating monthly trends to highlight seasonal patterns over time.
- Cluster plot (2011-2014): All four years are then shown simultaneously–one year per light strip–revealing year-to-year clustering patterns and overall variance.
- Following 2011-2014, the same sequence repeats for 2015–2018.
Red: > 4,000 cases
Orange: < 4,000 cases
Yellow: < 3,000 cases
Blue: < 2,000 cases
Final Thoughts & Takeaways
Concepting was one of the hardest stages. Many early ideas were either too simplistic, like basic bar charts, or too abstract to be legible. The final concept I landed on, a combination of a line graph animation followed by a cluster plot, struck the right balance. It allowed for variation over time while also giving a broader year-to-year comparison, all within the constraints of the four LED light strips under the El tracks.
Color and animation played a key role in enhancing comprehension: I used a color-coded system to communicate peaks in rat activity and accounted for viewers approaching from both sides of the installation. After initial tests, I refined the animation speed to allow more time for viewers to absorb the cluster plot and better connect it to the preceding line graph. A final prototype test with the lights revealed that some colors appeared too muted, leading me to adjust for higher saturation and visibility on the LEDs. This project strengthened my skills in both design thinking and storytelling through data and I really enjoyed the challenge.