Biometrics + Bus Stops: What Eye Tracking + Facial Expression Analysis Reveal

by Abigail C. Sekely + Ann Sussman, RA, theHapi.org

Figure 1. Eye-tracking Heatmaps of the Chelsea Bus stop, after and before redesign; the heatmaps glow reddest where people looked most. The original paired image, presented to viewers in study, appears on top. ©theHAPI.org

How do people look at bus stops? What catches their eye? What makes them happy? In this pilot study prepared for the City of Chelsea with ITDP-US (Institute for Transportation Development Policy), we asked these questions and learned just how much color and design matter when it comes to drawing people’s attention to a transit stop. 

The Chelsea Flower Walk project is a partnership between the City of Chelsea, BostonBRT, AdHoc Industries, Studioful Design, La Colaborativa, and local artists, that was launched in August 2023. Its goal: improving the bus transit experience, beginning with the very place riders wait by combining engaging graphic design, immersive art, plants and trees with a Zicla level-boarding bus platform to enhance accessibility.

The Research

As a key partner in the project, the HAPI team conducted studies of the before and after-art installations at the stop using biometric software from iMotions.com. Over the course of four weeks, fifty-nine participants took part in this 4-minute online study where they looked at paired images on their laptops (or desktops) with computer webcams tracking both their eye movements and facial expressions. 

Key Findings – Eye Tracking

Biometric tools, including eye tracking, are widely used in the advertising, automotive, and business world today to better understand the client experience. Eye tracking records both conscious and nonconscious eye movements revealing how we take in the world around us as the primates we are – often without conscious awareness or control. Figure 1, above, for example, shows the results of eye tracking images of the original bus stop paired with its redesign.

Eye tracking generates heatmaps that aggregate viewing results, glowing reddest where participants looked most, then fading to yellow, then green, as attention decreases, and finally, displaying no overlay color at all on areas ignored. Note how the heatmap, above, in Figure 1, glows reddest on and around the bench and then fades to yellow and green around the colorful flower patterns at its base; note also how there is no comparable focal point on the original bus stop – no red or yellow appear anywhere.

These results suggest how the original bus stop is unlikely to draw the attention of passersby. Eye-tracking studies of the built environment frequently reveal similar results; humans are hardwired to seek out organized complexity and shun blankness, no matter where we are or what we look at.

Figure 2. Heatmaps fall more on the new bus platform with flowers (left) than the original roadway (right). Top image displays the original pairing viewers looked at in the eye-tracking study. ©theHAPI.org

Figure 2, above, shows the power of the colorful flower patterns around the redesigned bus stop to capture attention. Here an image of the original blank asphalt road by the bus stop was paired with the newly designed Zicla bus platform. The heatmaps show how the new platform instantly drew viewers in, while the blank roadway could not, barely rendering a single bright red dot. The results also indicate the brain’s propensity to seek out edge conditions and areas of contrast; note heatmaps focused on distant trees and areas of color contrast; (this makes sense when you think about it, our perceptual systems evolved for survival.)

Key Findings – Emotional Experience

But how did the redesigned bus stop make people feel? Eye tracking records where people look, not how they feel, or emotional valence, which may range from positive to negative, to neutral. To capture how people felt about bus stops, the study included a preference questionnaire (see Figure 3) and used software that analyzes emotional expression bundled in with the eye-tracking software (from iMotions).

Figure 3. Preference questionnaire included in the bus stop  biometric study. ©theHAPI.org

Figure 3 shows how the preference question appeared in the study. At the study’s conclusion, we learned that 49 of the 59 participants (or 83%) preferred the new design, 4 participants (7%) preferred the original, and 6 (10%) indicated no preference at all.

Emotional experience may be further measured with emotional expression software (Affectiva), which tracks facial muscle movements and how these change depending on stimuli. While still photographs generally do not elicit a great deal of emotional expression, (videos generally yield stronger results) we found the facial data collected nevertheless correlated with the preference study. For instance, Figure 4 shows participants experiencing joy as they took in the redesigned bus stop. The red vertical line, below the heatmap, indicates where participants looked as they experienced positive engagement, showing views falling mostly around the redesigned stop.

Figure 4. Affectiva data creates graphs tracking Joy, (in green) Positive response, (dark brown) and Engagement overall, (brown). The red vertical line below the heatmap reveals where participants looked as joy spikes. ©theHAPI.org

Figure 5. Graphs track Joy, (in green) Positive response, (dark brown) and Engagement overall, (brown) as viewers take in the original and redesigned roadway. The red vertical line below the heatmap reveals where participants looked as joy spikes. ©theHAPI.org

Similarly, Figure 5 shows the new flower walk generating joy; note the heatmaps, falling on the painted walkway as positive engagement spikes.

Emotional heatmaps, in Figure 6 and Figure 7, below, provide yet another new metric for viewing emotional experience, presenting results at a glance. These use color washes over a study image to indicate which emotions are experienced where. Figure 6 shows joy and positive feelings surrounding the new bus stop, and Figure 7 shows positive feelings predominantly around the new flowery walkway (images presented in black and white so that the color wash can be readily seen).

Figure 6. Emotional Heatmap showing joy and positive response falling on the redesigned bus stop with virtually no glow on the ‘before’ bus stop.  ©theHAPI.org
Figure 7. Emotional Heatmap showing joy and positive response with a pink and red glow, primarily on the new bus platform and the planters in the ‘after’ photo (left).  ©theHAPI.org

Key Take-aways

The results here indicate how the 21st-century is a remarkable time giving us access to powerful tools, such as eye tracking and facial expression analysis, that can be used to reveal how the human experience of place happens. In the Chelsea study, we saw how people’s feelings and behavior change when blank surfaces around a bus stop are redesigned with colorful patterns, plants and flower art. 

This pilot study demonstrates how nonconscious behaviors direct the human experience in urban environments much more than most realize and current transit planning practices acknowledge. It also suggests the potential to design happier urban spaces by acknowledging and respecting our innate biological biases. This study specifically revealed:

  • How the original bus stop did not attract attention and will likely be ignored by passersby;
  • How adding color, nature and art to the roadway around the bus stop can change human perception and produce measurably positive experiences, including joy;
  • How quickly the engagement with a transit stop does – or does not –  happen;
  • How surrounding advertisements, street lights, and signage, tend to draw attention away from the bus stop rather than at it.

The study also suggests a path forward for future research, looking not only at transit stop design but also at roadway design and wayfinding, one that heeds human perceptual biology and biases. Given that humans tend to walk with their head looking down, (at an angle of about 10°-15°) how much does patterning on roads and sidewalks matter? Should we measure human unconscious experience in our transit plans? Should sidewalk and roadway design begin with it?

Today’s biometric tools place us at an inflection point; we now have a remarkable opportunity to understand humans better than ever before. The question becomes, of course, – will we use these technologies to build better places for people – or not. And if not  – why?

“The broader one’s understanding of the human experience, the better design we will have,” Steve Jobs once famously said. That kind of thinking is at the root of Apple’s stratospheric success; imagine applying it to create our public spaces!

###

For more on this story, check out the Streetsblog Mass article here;

Riders board the 116 from the “Flower Walk” art installation at the Broadway and 3rd Street bus stop. Photo by Ad Hoc Industries.

Further Resources on Using Biometrics in Design:

Adding Flowers to a Bus Stop Matters – Here’s Why!

Using Eye Tracking to Reveal Responses to the Built Environment and Its Constituents

Α Survey-based Study of Organized Complexity (found in nature and pre-modern architecture) and aesthetic preference

?s: email Contact(at)theHapi.org

Many thanks to City of Chelsea and creative partners on this project, including ITDP, Adhoc Industries, Studioful Design, Civic Space Collaborative, who made it possible; many thanks to iMotions team for their guidance and software + to the HAPI Team researchers Hernan Rosas and Alexandros Lavdas for their expertise.

This entry was posted in Uncategorized. Bookmark the permalink.

Leave a comment