An Overview

For this project, we started with a problem personally relevant to multiple members of our group, including myself: the harms and dangers associated with drunk driving. We then went through several rounds of brainstorming, user research, data analysis, and iterative sketching until we came to a finalized design plan and began designing our product. When our first version of our high-fidelity prototype was complete, we completed multiple rounds of user testing and design modification. To experience an interactive version of the final high fidelity prototype, see the bottom of this page.

‍

Our group won best overall design among our class of over 200 students and received a final grade of 100.7% on the semester-long project (a rare feat at Cornell University!). We also received runner-up for the social-impact award.

‍

The Problem

As is considered best practice for the UX Design process, we started with a "problem-first" mindset for design and ideation, rather than beginning the project with a final product in mind. We  brainstormed multiple problems that could benefit from a User Experience Design solution, and our group narrowed down which idea was both most practical and important to tackle.

‍

Below illustrates our group's topic ideas to illustrate the broad range of problems we were considering.

The problem that my group selected is the one I was most passionate about: the complex issue of drunk driving. According to the National Highway Traffic Safety Administration, every day about 37 people are killed due to drunk driving accidents in the US alone (about 1 person every 39 minutes). This alarming statistic is hard to understand when, for our group members, drunk driving is so frowned upon across our families and friends  and something that we never consider.

‍

We considered many different reasons why there may be such a high presence of drunk drivers despite so much evidence showing its risks. Our main conclusions were that these statistics are largely a result of two main groups: 1. Individuals with alcohol abuse problems but are still high functioning and thus drive themselves, and 2. Individuals who are not educated on BAC levels and who think they are okay to drive, but if they used a breathalyzer or other form of measurement, would realize they are not.

‍

The Solution

A product that prevents users from driving while intoxicated through combined physical and digital design components:

• A user-friendly application that displays measured BAC levels, tracks BAC history, alerts friends when drinking, and allows others to locate you even if your phone dies.
• A physical car key with a breathalyzer integrated into it for convenient access to BrAC (breath-alcohol content) measurements
• A wearable band that measures TAC (transdermal alcohol content) to cross-check the BrAC reading
• A feature that triggers an IID (ignition interlock device) in your vehicle when the BrAC and BAC measurements are too high
• An override method involving a cognitive test difficult for intoxicated individuals to determine if the BAC estimates were an error or if the individual is truly intoxicated

To get to the above solution from our initial problem, we conducted extensive user research and data analysis, which I will walk through next.

‍

User Research:

The first step in our user research was contextual user interviews. As a group, we brainstormed as many questions as we could  that we thought were relevant to the topic, and then narrowed down and categorized them into subjects. The final questions we used for our interview protocol are shown below:

We recruited eight participants that fit our user description of "current Cornell undergraduate and graduate students who identify as individuals that go to social events involving alcohol." We also chose a few interviewees who have their own cars on campus. Part of our interview protocol was ensuring we had our interviewee's informed consent, and explained that the interviews were completely anonymous and only identifiable by an interviewee number and not traceable back to them. This was an important consideration for us considering the sensitivity of the topic at hand. We had to ensure that interviewees would be willing to be honest if they had ever driven drunk in the past, which would be less likely if their answers were not anonymous.

‍

Data Analysis:

We created an affinity diagram of each datapoint from our 8 interviews. Datapoints included any insights we received from interviews, from demographic facts like participants' age, to their personal safety strategies for preventing drunk driving. After collecting the datapoints, we organized them into related topics. From these affinity diagrams, we distilled five main insights that would inform our design, which I included after the affinity diagrams below.

The five insights we gained from our user interviews:

Persona

From our user research and data analysis, we created a user persona, a fictional character created to represent a user who might use our product. We defined her goals, tasks, needs, opportunities, motivations, and frustrations, and used this persona throughout the design process.

‍

Designing

Once our user research, data analysis, and persona were complete, we began the design process. We used our user research to prioritize the datapoints collected about interviewees' ideas for solutions and strategies they use. We created 20 iterative sketches of potential solutions. We then created an affinity diagram with the best factors of our individual designs, and then voted on which of these features we wanted to include in the final design.

‍

Below are highlights of my 20 iterative sketches:

Below is an affinity diagram of the best features in our individual designs, which we narrowed down to finalize our design plan. We decided on a three component design system with a physical car key with an embedded breathalyzer and location tracking function, a mobile app, and a wearable device with apple watch integration.

Sketches

From these chosen design features, we created sketches of the three part design:

From our sketches and user research, we decided on 5 main tasks that we wanted our final design to accomplish, and sketched out scenarios to show the use of each of these:

1. Share real-time GPS coordinates with emergency contacts
2. Portable breathalyzer
3. Track BAC level history
4. Multi-device support
5. Cognitive test option to override IID if needed

To see further descriptions of the tasks and scenarios as well as their associated sketches, see this link.

‍

Prototyping

After sketching our designs and choosing design concept goals, we moved onto the next phase of prototyping.

‍

Paper Prototypes

We started with paper prototypes for their cost-effectiveness and ease of iteration with this type of prototype. Paper prototypes are ideal in these early stages of designing as, while higher fidelity prototypes look better and more similar to the final product, they take significantly longer to develop. A key tenet of a good designer is willingness to modify the design, which is often harder after spending so much time on a design and becoming attached to it.

‍

Below are images of my paper prototype, as well as a video walking through the prototype and each of the tasks we specified as milestones.

‍

Digital Prototype

After comparing paper prototypes, we chose our favorite elements of each. We were finally ready to start developing our high fidelity digital prototype.

‍

Below are images of our original digital prototype design:

‍

Evaluation and User Testing

Once our digital prototype was complete, we entered the "final" stage of the UX design process: evaluation. We evaluated our design in two categories: Usability Heuristics and User Experience.

‍

Heuristic Testing

We based our testing on Nielsen's 10 Usability Heuristics, shown below:

To test our design for heuristic violations, we each went through every page and feature of our prototype and noted any heuristic violations. These findings were compiled into the following spreadsheet: Heuristic Violations

‍

Usability Testing

To test our design for usability, we each recruited user testing participants that fell within our user group, and created a list of tasks we wanted them to complete. Examples of tasks were:

• View skin test history on Apple Watch
• Simulate a failed breath test and try to override it with the cognitive test
• View your friend Suzy’s location, most recent BAC measurement, and battery level
• Call an Uber for Suzy
• Check your highest BAC measurement on December 9th

With our test users' consent, we filmed them attempting these tasks, and took notes on any problems they faced. We categorized these problems by severity as well as type of usability violation according to the usability metrics shown below, and organized our findings in this spreadsheet: Usability Testing Results

From all of the above findings on heuristic and usability testing issues, we chose the most important problems to address and created a final prototype design. We thought our prototype was strong, and these modifications made the user experience even smoother and more intuitive.

‍

Final High Fidelity Prototype

The final high fidelity prototype Figma file is embedded below, followed by an interactive prototype you can try for yourself.

Conclusion and Reflection

This project was a long process with hard work that paid off. Because of the importance of the issue we addressed, as well as how strong our final design was, I am extremely proud of the end product.

‍

On our last day of class for the semester, we had a final project presentation where each team presented their projects on posters for classmates, professors, TA's, and other interested passerbys. Everyone we presented to was impressed by our project as evidenced by our winning the award for Best Design Overall. In a 200+ student class of UX design students, this was incredibly exciting and made us very proud.

‍

Our poster for the presentation: 

‍

What's next?

Because of the strength of our design and its significant potential impact, we are interested in expanding this project and eventually turning the prototype into a real app. There are many steps between the current stage and an actual marketable product, but achieving that one day is very exciting and motivating for us!

‍

Any future updates on this project will be reflected on this page of my portfolio.