Sustainable carpooling at GT

Team & Role

Led the ideation and design of key concepts, worked in parallel with another UX designer and 2 UX Researchers.

What I did

Ideated 13 concepts, sketched 3 shortlisted concepts, and created a storyboard for the final concept. Designed the application dashboard, carpool management and onboarding experience. Conducted 1 user interview and 1 usability test.

What I delivered

Analysis of Quantitative data, Research findings, High fidelity prototype

Project Duration

4 months

Discover, match, maintain & save trees.

Considering how the environment around us continues to deteriorate with every passing day, there needs to be a lot of work done in terms of creating sustainable solutions. The youth today realizes this more than any generation, and our project looked to leverage that by exploring a sustainable solution for commuting to campus.

Problem space

How can we encourage GT students to utilize carpooling services?

Carpooling permits sold at GT since 2018

Carpooling is one of the more well-known solutions for reducing individual carbon footprints and ensuring a better tomorrow. Yet, it is not very popular in current times. We decided to dive into this topic to find if we could do something substantial with carpooling in GT. When we found out that GT already had a carpool system in place, we endeavored to find out why it was not more visible.

Gathering Context

Understanding the various stakeholders

When referring to carpooling in Atlanta, there is a broad set of stakeholders, particularly when accounting for all factors within the socio-technical context: 

At the core, are Georgia Tech’s students, faculty, and staff. Expanding further, our internal stakeholders are Georgia Tech’s Department of Parking and Transportation as well as the office of President Dr. Ángel Cabrera. External stakeholders include the City of Atlanta, whose department of city released the latest Comprehensive Development Plan (CDP) on August 27, 2021. Finally, general commuters (others who drive on the streets of Atlanta) and neighboring communities such as Home Park account for the public stakeholders.

Narrowing down our user group

After studying and talking to various stakeholders involved with carpooling at Georgia Tech, we narrowed down on our user population to graduate students of Georgia Tech who are eligible for carpooling. What are their characteristics?

What are the characteristics of our user group?

Task Analysis

How would a member of our user group access carpool permits currently? 

User Research

We gathered user data through 6 interviews, commuter surveys and google analytics.

Semi-structured interviews

Being less structured, our approach offers a conversational tone and affords a more casual dialog and opportunities to “open up”, rather than a unidirectional communication flow. By creating this gateway, we were able to touch on topics that may be considered sensitive to some, such as issues of equity, diversity, and socio-economic standing.

Commuter Surveys

By using the surveys, we were able to gain insight into the general Georgia Tech population: faculty, staff, and students. Due to the sheer number of participants, the information collected offers a better description of the respective attributes of the general population.

Google Analytics

Using this platform, we can determine how users are arriving at the site, where they visit, where they leave, what devices they use, and how long they’re engaged.

Analyzing Quantitative Data

Number of responses

Our team obtained data from the annual Georgia Tech Commuter survey, an annually distributed survey by the Parking and Transportation Services (PTS) for all registered commuters.

Key Finding #1

60% of commuters consciously chose not to carpool

Q9 of the survey
Analyzing the behavior flow
Key Finding #2

No one who looked up how to apply for a carpooling permit applied for a permit

In the behavior flow diagram we see that 2200 students open the carpool rideshare page to look up initial information about carpooling. Out of these, 148 people went ahead and checked on how they could apply for a carpool permit. Out of 148 users, 0 users clicked on this link to begin with their application. This further confirms the response we received from a group of our users who mentioned that the process appeared to be too complicated and prevented them from applying for a carpool permit.

Key Finding #3

17% of the users drove alone because they didn't have a carpool mate

Not having anyone to carpool with and not wanting to depend on others were the two main obstacles towards carpooling.

Analyzing Qualitative Data

The interview protocol consisted of 11 primary questions with varying probing questions. Each interview session lasted between 20 - 30 minutes, and was semi-structured.

Key Finding #1

Users need help while forming a carpool

Through both our survey as well as our interview, we discovered that many users are not comfortable forming a carpool by themselves and require assistance during different stages of carpool formation.

These complaints by users point us to a very important inhibitor to carpooling which was highlighted in majority of the survey responses.

Key Finding #2

Users want their carpool riders to have compatible schedules

From both our surveys and interviews, we received a large number of responses suggesting that users cannot opt for carpooling because their schedules do not match with people that they would consider carpooling with.

Key Finding #3

Users would need help finding a suitable parking lot

This is an intervention we hadn’t anticipated beforehand, but we learnt that it is equally important that carpoolers choose the right parking lot in order to ensure there isn't a large disparity between the amount students have to travel to reach to the car.

Key Finding #4

Users want their carpool mates to have similar interests

In contrast with cab services such as Uber and Lyft, carpooling is often seen as a social experience. This expectation is seen across responses from our user interviews and surveys, where users have indicated to us that they want carpooling to be a good social experience for them.

Key Finding #6

Users were concerned about their safety while carpooling

Safety is a major concern when you are regularly riding with people you are not too familiar with. In our interview and survey students indicate that building trust and a sense of safety is a big deciding factor for them.

Design Requirements:

  1. Simplicity: A large number of students have stressed on having an experience that is straightforward and simple. The system should hide the complexity from the users and render a simple interface on the front-end.
  2. Accountability: The system should offer a method of accountability to report repeated violators and have a clearly defined escalation plan to address the situation. 
  3. Discovery: We learnt that users have immediate obstacles to find potential carpool matches. If the system does not return acceptable matches based on their prioritized criteria, users may abandon the carpool process entirely, regardless of the interface’s usability.
  4. Sustainability: The ecological impact of carpooling is a mild incentive to potential carpoolers; however, by elucidating their environmental contributions, we may galvanize their proclivity to choose carpooling in other contexts.
  5. Safety: One of the challenges we discovered in D1 was the concern of safety and carpooling, particularly with students. The system should have procedures in place that would prevent malefactors from abusing the data contained within the database.


I ideated 13 design ideas, created concept sketches for 3 shortlisted ideas, and a storyboard for 1 out of the 2 final ideas.

In order to help us break down each step of the user process and analyze possible design interventions, I and David used the Slice and Dice strategy to analyze Georgia Tech's carpooling service while Linda and Sejal analyzed the carpooling feature in Waze.

We then collectively conducted a task analysis for the PTS website, that further went into the user process while signing up for a carpooling permit.

SCAMPER brainstorming

We further made use of S.C.A.M.P.E.R. brainstorming strategy. By restricting my thinking to specific functions such as modifying existing solutions, or eliminating features, I was able to think with more clarity and direction. As a result, I was able to contribute several suggestions aimed at improving the experience current carpooling options offer and improving the way information is conveyed through Georgia Tech's carpooling page.

From our research responses it seemed like people had various preferences about viewing data. One group of users knew what they wanted and were looking for specific country-related information. On the other hand, the other group of users were exploring and browsing through trends in their countries as well as other countries.

The design solutions I proposed were centered around using visualizations to create an engaging and user-friendly user experience.

Since a lot of our interviewees mentioned a matching schedule is of utmost importance to them, I felt that we could easily calculate the similarity between 2 schedules and display the score as a percentage. This would allow users to swipe based on how important it is to them.

Another feature that I strongly felt for was the parking lot optimization feature. As one of our participants mentioned in the interview, choosing the optimal parking lot can be an issue when more people are involved. Using geo visualizations our application could offer a design intervention here and calculate a parking zone that is at the most optimal distance from the class buildings of all carpool members.  

Finally, given that sustainability is an important benefit of carpooling, I felt it was necessary to visualize how the users' choices are positively impacting the environment in order to give them a sense of achievement.  

Concept #1: I Am Carpooler

I Am Carpooler shows visualization to users, where the design indicates how much money they are saving, relation to sustainability, gas savings, route for carpooling, and expected time for leaving and heading back home. The visibility feature designates information into a screen where users can easily recognize and understand the information. 

Requirements satisfied: simplicity, sustainability, discovery

Concept #2: Buzzpool

Buzzpool is a design concept that helps users to find potential carpool partners. Users can upload one’s information by logging into GT login feature, which enhances security and provides users with a sense of comfort. Based on the user's preference, the application provides the list of potential carpool partners, indicating bio and matched percentage. 

Requirements satisfied: safety, matching schedule, helps matching potential carpool mate, limits audience to only GT students

Concept #3: Our Lot

Our Lot is an optimization concept that allows users to easily distinguish which parking lot is suitable for the carpool group. The feature will indicate an optimized parking lot zone based on the carpool members’ schedules and locations of classes. Our Lot shows an intuitive design that provides convenience to users because they don’t have to compare locations and times and try to find a lot that meets their needs.

Requirements satisfied: proximity, simplicity, discovery

Storyboard: Buzzpool

Strengths & Weakness:

One of the key strengths of my design was that it addressed user concerns of safety, flexibility and discovery. Firstly, buzzpool ensures the safety of the user as it is only limited to graduate students and operates on GT servers and information. Buzzpool also allows users to choose carpool mates based on various parameters like gender, location, music, schedule etc. This allows the user to require more accurate results about potential carpool mates.

A weakness of Buzzpool is that initially, it may seem similar to existing carpool applications like Waze. However, since Buzzpool is only for GT graduate students and gives users more information about potential carpool mates, we plan to highlight these features to showcase the differences from existing systems. Additionally, since the application is restricted to only our user group, i.e., graduate students at GT, it is possible that users may have a smaller pool of applicants to choose from.


After recieving client and developer feedback on my concepts, we designed mid-fidelity prototypes focusing on presenting data through the world map, country dashboard, comparison view.


The onboarding screens were designed by me. The purpose of the various questions is to prompt users to input their preferences, which would be used to find their potential matches. The process consists of 8 steps ranging from driver/rider preferences, schedules, and music genres. In pursuit of Nielsen’s #1 usability heuristic, I created a progress bar to inform users of their stage within the process. I employed the use of a Ford Model T and traffic lights to resonate with the Georgia Tech spirit mark, the Ramblin Wreck.

Loading Screen

In continuation of providing the system statuses, we created a load screen in between the onboarding screens and the match results. In regard to principles within the Doherty Threshold, we embedded pseudo quotes from famous musicians to visually engage our users while the search queries were occurring in the background. We found these quotes offered a reprieve and comic relief during our users’ wait time.

Discover Matches

Our discovery screens allow users to view potential matches that they can consider for carpooling. The matches are sorted based on the user’s preferences.


In an effort to highlight sustainability, users are informed of their environmental impact using tree consumption as a metric. This metric relies on the user’s existing commute plan. As they become more environmentally friendly, their status will change to one that is saving trees.


The Inbox screen can be accessed from the dashboard. It allows the user to view all new matches and have conversations with other users.

Carpool Management

The My Carpool screen gives the user all the information pertinent to their carpool. It shows all the carpool members, the daily schedule of the user, as well as provides them with options for Parking, Route and Contract. The Contract option builds-in an accountability feature. This feature allows users to have a voice in their carpool, and holds each member accountable for their actions, thereby reducing conflicts and making the carpool a much more pleasant experience.


The Route option allows users to view the most optimized route for them to take to come to campus. It allows users to view the quickest route for them, thereby reducing anxieties caused by delays - one of the many deterrents we had observed in our research. In our prototype, the final user interaction is viewing the carpool’s route for arrival and departure.


We interviewed 4 users and gathered quantitative data on the SUS scale.

Our test sessions occurred simultaneously virtually over Teams. The sessions were conducted 1-on-1 and recorded. Each session took approximately 30 minutes. The general approach was to have each participant think aloud as they accomplished 3 assigned tasks. After each task, we asked them about the ease or difficulty based on a 5-point likert scale. We also asked post-task questions to elicit qualitative data about their experience with the system. Near the end, each participant completed a System Usability Scale (SUS), followed by closing remarks and feedback.

In order to facilitate participant interactions with the design prototype, we created a persona of Susie Burdell. The participants were then presented with tasks that they had to complete as Susie.

Task 1: Create a profile on Buzzpool for Susie

Task 2: Create a profile for Susie

Task 3: Create a carpool contract for Susie

Key Finding #1

Users enjoy viewing their environmental impact in an illustrative way.

2/4 users responded positively to the illustration that highlighted the amount of trees they were saving. Users were able to notice the change in their dashboard illustration without us having to point it out.

Key Finding #2

The system informs users of their status in the onboarding cycle

Users also wanted a persistent system status (sticky progress bar) during the profile creation process. In general, users wanted larger interaction footprints (Fitt’s Law), where they could not only click/tap on interactive elements such as radio buttons, but their labels as well.

Key Finding #3

By using similar UI patterns from dating apps, users may bring with them conflicting expectations

Users pointed out that the matching system reminded them of a dating application. They expected to see real pictures, and swipe interactions to find matches.

Key Finding #4

Users expect further validation when adding carpool mates

While adding a carpool member, a user highlighted that they would prefer to see a dialog box that notified them that their interaction was successful.


Reducing the onboarding sections

I reduced the amount of sections our onboarding experience was divided into. I felt that having so many sections shown to the user in the progress bar could potentially overwhelm them and place a significant amount of cognitive load.

Instead of having one question take up one screen, I decided to group them into 3 based on their theme and allow the user to scroll through multiple questions of the same category.

Making the avatars look livelier

Some users complained that they would like to see real pictures. In Phase 4, we intentionally did not put real pictures in order to safeguard the privacy of our users and distinguish ourselves from dating applications.

As a result, in order to prevent user profiles from looking empty, I redesigned the avatars in order to give it a more lively look but at the same maintain a degree of professionalism.

Users complained that they would like to see some sort of a confirmation when they send someone a carpooling request and when someone accepts it. In order to address that I added these dialog boxes.

Using states to guide users to the next step

Some users said that the My Carpool screen can be a little overwhelming at first. As a result, I redesigned the intial state of these screens in order to better guide the users through the process of carpooling

Allowing users to edit rules, and making the contract UI more realistic

Users were a little confused by the contracts process earlier on due to the lack of better design that reflected a finished contract. They also said they might want to make changes to the contract but we did not have an option to revise it.

In the new design above, I give the contracts a more realistic touch, and at the same time allowing users to revise the version and resign it.

I changed the styling of the parking screen in order to make it more consistent with our application style.

I redesigned the route path style to be more consistent with how it is in google maps since users would be more familiar with it.


Lessons Learned

Throughout the course, we became familiar with the iterative design process and learnt how to incorporate the skills learnt in the classroom into our project. Since each member of the team came from very diverse backgrounds, we all learnt different skills and techniques from each other. During the earlier phases of the project (D1 and D2), we learnt a lot about the problem space and possible solutions by exploring a multitude of external resources. Through this process we learnt how our design system could interact with a larger ecosystem, like the City of Atlanta’s plan for reducing parking spaces. Additionally, each of us strengthened our research skills through the project. We learnt how to extract key findings from each phase of research and translate these findings into meaningful design iterations that are backed by data.

What we would do differently

Looking back on our project, there are some things that we would want to change or improve upon. Firstly, we would like to go back to our drawing board and find more ways of integrating sustainability into our design prototype so that it was more central and integral to the prototype. We would also want to get feedback from our user group during our earlier iterations or wireframes. If possible, we would have liked to engage with our user population and include them in the design process to ensure that our prototype successfully addressed concerns of accountability and discoverability. 

Having user feedback during initial iterations would have also helped us test whether the app matched the mental model of our participants or not. Something that was a prominent concern during our discount evaluation. While we appreciate the feedback provided by our peers, we would have also liked to present our prototype to our user population and get their feedback. Overall, we learnt a lot during the project.