Mockup of final Multilingual Templates feature

Rainbow Road: Redesigning the Escalator


Long story short
We studied transition spaces using two kinds of research methodologies: the traditional scientific method and ethnographic methods. Afterward, we created a redesign (The Rainbow Road 🌈) based on our findings.
ROLE: UX Researcher, Designer, Project Lead
Chris Sun, Takeria Blunt, Xu Zeng
Fly-on-the-Wall Observation, Card Sorting, AEIOU, Participatory Observation, Experience Sampling, Directed Storytelling
Sept-Nov 2018

Project Description

What were we tasked with?
For this project, we studied transition spaces using two sets of complementary research methodologies:
The class split into groups of 6, each focused on a different transition space. Afterward, we split into groups of 3 so that each sub-group would use different methods to study the same transition space from two different lens.

Our group decided to study escalators, specifically the ones in our local Barnes & Noble on campus. We chose this as our transition space because not only is it a physical transition from one floor to another, but you frequently see changes in how people interact with the space around them as they get on an escalator. You can see them pull out their phones, look around the bookstore, or even just keep walking up. There’s a rich variety of interactions that occur on an escalator that would be interesting to observe and analyze further.
User flow diagram for adding translations

Georgia Tech’s Barnes & Noble, escalators on either side

There are three parts to this project:

Scientific Method Research


What do we do?
In order to get some ideas of what methods to use for the research methods rooted in scientific tradition, we looked through Bella Martin and Bruce Hanington’s Universal Methods of Design (2012). We selected some methods that we thought would be most effective in making observations that are reproducible, verifiable, and independent of the observers.

From there, we narrowed our choices down to our top three based on how well we thought they were rooted in the scientific tradition. We chose methods that were different enough from one another so that we could have a variety of perspectives and observations, but also worked well with one another to illustrate a bigger picture. We ultimately decided on Fly-on-the-Wall Observation, Card Sorting, and AEIOU, as they included some observational methods, as well as methods that could help us organize our observations to better synthesize our ideas. These methods were also reproducible and verifiable, which are important aspects in methods that are rooted in scientific tradition.
User flow diagram for adding translations

Our three chosen methods: Fly-on-the-wall, Card Sorting, and AEIOU

Initial Findings

Let's see here ...
Our sub-group prepared a brief interim presentation that discussed why we chose our transition space, our chosen methods, and what our initial findings were. Fly-on-the-Wall Observation allowed us to simply observe our transition space without any predefined criteria. It gave us a ‘secret outsider’ perspective that was ideal for a public space to make observations without disrupting or influencing those who we were observing.

Some of our initial findings included:
User flow diagram for adding translations

Fly-on-the-Wall Observation Notes

Card Sorting allowed us to organize our observations into three main aspects: getting on/off the escalator, movement on the escalator, and pastimes on the escalator. The most variety of observations we found were the interactions people did while on the escalator, such as tying their shoes, reading, and stretching.
User flow diagram for adding translations

Grouped observations through card sorting. Main groups in red, yellow, and blue, while intermediate groupings in orange and green.

AEIOU helped us be more focused on five specific aspects while making our observations: activities, environments, interactions, objects, and users. This was integral in deconstructing this transition space into objective, observable characteristics.
User flow diagram for adding translations

Our AEIOU method findings

While we went into this with no research question in mind, we did come out of it with several curiosities:

Final Findings

Let's try it again!
While we did not have a research question in mind for the previous iteration, we used one for our final observations: how does the direction of movement (going up vs. going down) affect behavior? We noticed in our interim observations that some behaviors occurred more often while people went up the escalators than down, so we decided to focus in on what interactions people did in each direction and what that could mean in relation to the direction they were going in.

Our main findings from our second iteration of Fly-on-the-Wall observations include:
User flow diagram for adding translations

Revisited Fly-on-the-Wall observations

In our Card Sorting, we grouped our observations by their interactions: walking/running, phone usage, and talkativeness. We also noted whether the interaction was going up or down the escalator. Our findings were similar to the Fly-on-the-Wall observations: users tended to be more active going upward.
User flow diagram for adding translations

Revisited card sorting groupings

Our AEIOU observations had no significant differences from the first iteration because the activities, environment, objects, and users were similar. The interactions were noted during the Fly-on-the-Wall observations.


What'd we end up on?
Overall, we saw more active interactivity (walking/running, talking) on the way up, while more passive interactions (phone usage) on the way down.

Our overall findings showed that there was more movement going up, possibly due to more intent in moving to the top floor (such as to find something) or with some unease with just leisurely riding upward. More phone usage was found while going downward, possibly because customers were in no rush to checkout. In both directions, however, we found there could be a social influence in how one goes about riding the escalator; we have noticed several times that when someone in front begins to walk, then the person behind will also begin to walk, as well.

Since these methods were primarily passive observations, these are merely assumptions that will be assessed in the next section of the project.

Ethnographic Research


How'd we pick the methods this time?
For the second part of this project, we went through pretty much the same process as the first part. We once again went through Bella Martin and Bruce Hanington’s Universal Methods of Design (2012) to find methods that would work best for ethnography.

After compiling our list, we once again picked the top three methods that we thought would be the most effective in learning and understanding the people of the “culture” we are studying, which in this case are escalator riders. We decided on Participatory Observation, Experience Sampling, and Directed Storytelling because we believed that these would give us the most rich data about what our participants are thinking about.
User flow diagram for adding translations

Our chosen ethnographic methods: Participatory Observation, Experience Sampling, and Directed Storytelling

Using the Methods

Enough talk - time to act!
We began with a research question: why do people hesitate while getting on/off escalators? We had noticed this in the first part of the project and thought that this would be interesting to look at and understand the sentiments of this frequently-observed pattern among escalator riders.

For Participatory Observation, we each rode the escalators up and down several times. We kept track of our feelings, thoughts, and emotions while riding the escalators to try to gain a more empathetic view of what typical riders would think of and do on an escalator. We also made sure to note what we were thinking while getting on and off the escalator.

For a study as specific and short-term as riding an escalator, it was difficult to use this method without being hyper-conscious of our actions, so most of our data was skewed.
User flow diagram for adding translations

Participatory Observation notes

Our Experience Sampling involved asking participants to tell us their thoughts and feelings at given intervals (around 10 sec, 20 sec, and 30 sec). The first and last few seconds gave us insight on their thoughts immediately after getting on and immediately before getting off the escalator.

At 10 seconds:
“What books should I get upstairs?”
“This escalator is so long.”
“What books did I miss downstairs?”
“I need to stand on the right side.”
“This escalator is so tall and I’m scared of heights.”
At 20 seconds:
“How much longer is this escalator ride?”
“Are the checkout lines worth standing in?”
“I should just start walking the rest of the way.”
At 30 seconds:
“I have to be careful getting off.”
“Where do I go now?”
In order to do Directed Storytelling, we asked customers that just got off the escalator to describe to us their experience riding the escalator. For this, we referred to Kathy Charmaz’s intensive interview style (Grounded Theory) to better invite reflections by the participants. This helped us request more detail about their thoughts, feelings, and actions while on the escalator.

During the escalator ride, participants seemed to have a variety of thoughts:
“I skim the bookshelves to find a book that might look interesting so I can check it out on the way out.”

“I think about which direction I need to turn once I reach the top.”

“This was my first time in this store so I was mostly looking around just to see what’s available here.”
However, many participants appeared to be more hesitant when getting on and off the escalator:
“I start paying attention to my feet toward the top because I got caught in an escalator when I was younger.”

“I hesitate before getting on the escalator for a bit because I’m afraid I won’t fully step on it if I’m not careful.”

“Getting onto the escalator is scarier than getting off of it.”

“Escalators just look kind of scary.”

Findings and Synthesis

What does all of this actually mean?
While it was fairly difficult to get deep insights about riding escalators, one interesting theme we  noticed was a common fear of getting on and off the escalator.

From our findings, we arrived at several conclusions:

Redesigning the Escalator

Project Description

Wait, I thought this was research? We're designing now?
This project focuses on framing a problem and devising a design intervention in response based on the methods described previously.

The scope and aim of the design intervention may include:
  • Addressing a problem to “improve” the transition space (interpreted widely with concert material consequences)
  • Probing further into the transition space to produce new insights
  • Challenging the underlying assumptions of the transition space
  • Critiquing the transition space as a technical or cultural phenomena through an artifact
For our design intervention, we chose to “improve” the transition space through a redesign based on our findings.


Back to the drawing board .. literally ..
Based on our prior ethnographic study of escalators, we pinpointed a fear of getting on and off the escalator as the primary issue. Focusing on this problem, we then began to brainstorm a list of possible redesign ideas that would help make the escalator ride a more positive and less fearful experience. At this point, we realized that these issues are also very applicable to handicapped individuals, such as wheelchair users. We then decided to include accessibility as one of our main focuses.
User flow diagram for adding translations

We brainstormed 100 possible redesign ideas just to get the juices flowing

While brainstorming, we shifted our primary focus to accessibility for handicapped riders, with a secondary and simultaneous focus on resolving the fear of getting on and off the escalator. We culled our ideas by feasibility (is this doable?) and effectiveness (does it solve the problems we want to solve?).
User flow diagram for adding translations

Some early sketches of how we could consider accessibility in our redesigns

Initial Prototypes

We went a little haywire ...
We created two different prototypes for our initial designs, the “Rainbow Road” and the “Magic Carpet”, each providing different solutions to the problems we wanted to address.

The Rainbow Road was our most feasible solution, the main design component being creating a larger step for accessibility purposes. This was based on a system that has already been implemented in some locations in East Asia, though our proposed solution would be entirely automatic to preserve the user’s autonomy.

Here’s how it would work:
The other main design change we made was coloring each step. This makes it more clear to riders about which step they’re stepping onto in order lessen their fear of missing a step or accidentally stepping between two steps when getting on the escalator.
User flow diagram for adding translations

The “Rainbow Road” physical prototype feat. Xu’s hand 👋

Our second design was called the Magic Carpet. This is because instead of having a continuous staircase like normal escalators, a “magic carpet” box is spawned whenever a rider wishes to move up or downstairs.

Here’s how it would work:
The design combines elevators with escalators in order to provide a safer experience without fear of falling backwards off a step. The individualized riding experience makes the rider more comfortable by having a remote that helps control their movement speed; this also alleviates their fear of trying to make it on to a moving step by having each “magic carpet” held still until the rider is ready.

Check out some of the concept sketches below.
Request for a feature to translate and allow multiple languages for templates with 357 votes on it
Request for a feature to translate and allow multiple languages for templates with 357 votes on it
Request for a feature to translate and allow multiple languages for templates with 357 votes on it

Concept sketches for the "Magic Carpet" escalator

Critiques and Feedback

What do the people think?
From our initial prototypes, one of the main critiques was that we need to be less conceptual and more feasible. We also needed to create a more accurate model that illustrated how the proposed design would actually move and how users would interact with it.

Some other considerations that came up included:

Focusing the Scope

Let's bring it in
As per our critiques, we focused our project statement to:
“Creating an accessible escalator that aids in moving larger, wheeled structures up and down stairs in a safe way without having to navigate to an elevator.”
By making an accessible escalator system, we help users with different needs stay within the “main space” (ex: a mall) without the stigma of having to travel to an elevator in another location. This also makes the escalator still accessible to regular users so that this new design is inclusive of almost every rider, rather than being exclusively for handicapped riders.

Final Prototype

Let's bring it in
Since the Rainbow Road was our most feasible idea, having already been engineered in parts of Asia, we decided to continue with that design.

Our main changes include:
We chose to keep the steps rainbow colored because according to a study by Urban Realities Lab and Happy City, participants crossing rainbow crosswalks implemented in Vancouver reported being 40% happier in the space than in regular crosswalks. By translating this concept to our escalators, we not only hope to clarify each step to the riders so they can step onto the escalator with confidence, but we also hope to enhance the escalator riding experience overall and build trust between the rider and the escalator. We intend for the steps to be lit up by LED lights so when steps link up together, then the joined steps become the same color to confirm to the rider that the steps have successfully linked together and clearly indicate to other riders that there is now one large step.

Here’s how the new Rainbow Road will work:
Check out the prototype videos below to see it in action!

Project Reflection

Reflection on Scientific Methods

Overall, these methods worked well together and provided an effective way of observation and analysis that allowed us to better understand escalators as a transition space from a purely observational standpoint.

Reflection on Scientific Methods

These three methods overall gave us a very holistic understanding of what people think when riding escalators, though they could have been more focused to be more effective in our specific context.

This project was quite informative and allowed me to finally learn about and apply research methods commonly used in industry. It also gave me a better understanding of how different methods can be more focused in different fields (scientific method vs. ethnography) and may fit one context better than another. Studying such a specific transition space made me really consider unique ways to approach applying these methods, as well. Being hands-on with these methods definitely helped me be a better researcher and helped me understand different research methodologies on a more personal level.

Reflection on Redesigns

This design effectively resolves many of the issues we found as a result from our previous project in a way that is accessible to new users while also improving the experience for previous users. While we were first solely focused on designing for regular users, this project really helped us understand how to design for new audiences and how to consider them when making redesigns.

Here’s some of my key takeaways:

In the Future

For future designs, we hope to be able to work out some of the other logistical issues in implementing the Rainbow Road design, such as raising and lowering the handrails as the platforms are being used. If we had more time, we would have done some evaluations and critiques with some handicapped individuals, as our evaluations were conducted with all able-bodied individuals. This would have given us a much better insight for the direction of our redesigns.