maddy cha

Safe & Sound

Increasing Safety on Campus


contextual inquiry


affinity diagramming

experience prototyping

ui/ux design


Candia Gu

Josh Kery

Amru Palaniyappan

Every semester CMU alerts its students about at least one incident of attempted assault, or chase, or sexual harassment targeted against a student in the university area. Safe & Sound helps to combat this and increases student safety through providing locations of other students on their route, speed-dialing/texting preset contacts, and scaring off potential threats.

I helped to conduct and analyze user research for each step of the process and was in charge of creating the application designs and supporting visuals.
Log-in Page
Secure log-in using CMU ID ensures that only CMU students may enter the community
Map & home page directs students to their destination and shows other nearby students on their route home (anonymous)
Alert Settings
Allows students to alter default settings including who and how to contact in case of emergency
Alert Button
Press and hold interaction allows for ease of use while decreasing chance of accidentally sending alert.

Problem Space

Over the course of a semester, we were tasked with thinking about CMU on-campus transportation and the needs of the student population.

Through our initial surveying and interviewing, we began to notice trends revolving student safety. For one, students felt more secure when walking home with others or walking near other groups of students.

Stakeholder Map

We looked into existing services at CMU that focus on student safety/transportation, namely the blue lights services, CMU police, and various popular transportations (Uber/Lyft, PAT Buses etc.).

Through creating a stakeholder map, we understood their relation to the various stakeholder in our problem area, such as the students themselves, the university administration as well as parents and those who cared about the students. By identifying the stakeholders and understanding these relations, we were able to understand the context we were working in as well as the target audiences we had address.

Contextual Inquiries

We conducted 4 contextual inquiries with undergraduate students. Students told us whether the lived on or off campus, why they chose to live there, how they got to and from campus, during what type of situations and under what conditions they felt safe or unsafe, actions they took to feel safer, and even highlighted specific experiences where they felt unsafe.

The most common responses, for which specific notations are provided below, were that students felt safer when they were walking in a group or when there were more people on the streets and that calling CMU police wasn’t easily accessible.

Affinity Diagram

We consolidated all of the data gained from the contextual inquiries to create an affinity diagram. We grouped based on key ideas and implications from each of our inquiry notes.


For the survey, we created the questions based on the general statistical questions we wanted answered. We then distributed them to various Facebook groups as well as individually sending them to some of our peers. We received around 50 participants with a focus on undergraduate students at CMU.

Over 65% of respondents answered that they live relatively close to or on campus and, as a result, most students also walk home from campus. More than half of our participants responded that they walk home late at night between the 10pm and the morning. Additionally., we found quite a lot of variability with students’ knowledge of existing safety services provided by CMU.

Walk the Wall

We consolidated all of the data we had collected up to this point from our contextual inquiries, affinity diagram, and survey results. By looking at all of the information altogether, we were able to narrow down our problem space.

Throughout all of these generative methods, we found that a majority of students tend to walk home late at night, and rated safety as a high priority in comparison to other aspects of transportation (i.e. money, time, distance). However, participants of our contextual inquiries all seemed to think of their safety retroactively, worrying during their walk home rather than taking preparative measures prior to going out to be more secure.


From our walk the wall activity, we were able to generate a couple user needs that we could test using storyboards. For this part of the process, I decided to focus on the need for a sense of security through company, and created different solutions of various risk levels—focusing more on the need and less on the technology of the proposed solution.


We tested with a variety of undergraduate students who walked home as a way to validate the needs the storyboards were based of.

We found that generally the needs that we investigated where present among our target users. The solutions the user usually preferred were the ones that were the most convenient (i.e. having a phone application rather than having to carry other/new equipment).

From what we learned from speed-dating, we revised a few of our ideas and voted on the idea(s) we thought might be successful and efficiently implementable.

Paper Prototype

After identifying which needs and ideas were reinforced through our storyboards and speed-dating activity, we noted all of the key features we wanted to include.

Based on a combination of wanting to feel secure and wanting to have people on the line or texting to act as “witnesses”. We worked together to create an application that would allow users to see that there are other people around them while also allowing them to alert specific contacts if something were to go wrong.

Mid-Fi Iterations

As most confusion from our experience prototyping stemmed from the fidelity of our prototype, our main focus was making sure that our iconography and screens were more comprehensible. From there, we also worked on specificities in terms of our alert preferences and how exactly the alarm would look and function.

We decided to go with a dark themed background so that users could keep the application running without losing excessive amounts of battery while making it less distracting at night.

Experience Prototyping

With our mid-fi prototype, we tested walking home with the people from our speed-dating activity to see how they would utilize the application.

Commonalities between users were the following: users found that the emergency alert button on the map screen was in a very prominent location, easily accessible (and maybe too easily accessible). One user commented that the button might be placed to similarly to a home button

Additionally, users wanted more granular options for calling and texting different contacts (i.e being able to choose to send a text to their parents but call the campus police).

Final Design

Taking into account the user feedback we received from the experience prototyping, we continued to iterate on the design.

We decided that we would want to limit the application to the CMU community (through secure login with CMU ID) to reduce the risk of abuse of the application, especially considering the live location services. Additionally, we worked to alter the placement and visuals of the alert button link so as to reduce confusion with that aspect.

We also made it so that users could choose different methods of communication for different contacts if they so chose.


With this project, I gained a much more in-depth understanding of a variety of different evaluative and generative research methods and when/where in the process they are the most effective. One of the biggest take-aways I gained was to always keep an open mind, especially to new ideas. Being able to extract insights from user testing to further iterate and improve on our idea was an incredibly rewarding experience, especially with such a salient problem space.