CONVOY is a wearable device and an app for women to feel safe walking at night 

Product Design, Wearable Device Design, UX Research, UX/UI Design, Hardware UX


August 2019 - December 2019



Figma, Rhinoceros, KeyShot, Photoshop, Illustrator, 3D Printing, Arduino


Anjali Devakumar, Morgan Chin and Yujin Xue

iPhone 12 Pro.png

The Problem

In the United States, 45% of women report they do not feel safe walking alone at night

The Solution

An app and a wearable device that suggests safe routes according to crime data of an area.

The wearable device guides users to the desired destination by using voice and haptic interaction. Users can get information about neighborhoods and receive the latest crime and news stories.



01. Background

Literature Review and Competitive Analysis methods were selected to explore the problem space.


Literature Review

Socioeconomic status, ethnicity, and age can increase the levels of anxiety experienced when walking alone at night.
Women tend to arrange their plans to occur in daylight, walk different routes, and do not attend events at night to stay safe.
They feel the urge to let their friends and family know when they arrived at the desired location safely.

Competitive Analysis

Despite the high statistics and adverse effects of this problem, a robust product or a system does not currently exist. 3 crucial issues with these products are defined:



Lack of feedback, state of the product, and communication.



The interaction is either takes a lot of time or the language is too complex to understand.


Error Tolerance

Once the emergency state is activated it can't be canceled.

02. User Research


User Research

In order to understand our user group better, we have conducted various user research methods such as surveys and semi-structured interviews + follow-up questionnaires.



36 participants

Frame 3.png

~ 15 mins


18 questions

To understand the trend of what our users make them feel safe, we distributed a survey. This helped us to categorize certain behaviors and perceptions towards walking alone as women.





9 female participants

Frame 3.png

~ 1 hour


20 defined questions

Statements of our participants bolstered the information about not feeling safe walking at night. 

Quality of the environment (maintenance, lumination, etc.) that provides safety changes according to the individual.

In contrast to our assumptions, users don't want to carry items or weapons in their hands and don't prefer to be distracted by their phones.

Participants's own handwritings about what makes them safe or not

03. Understanding Data


Understanding Data

To understand our data, we created an Affinity Map with the data obtained from user research activities.

Pain Points

Pain Points

We found 19 major pain points that all were unique in their situation, contexts, and relevant to the original problem. We decided that there were six insights that we wanted to concentrate on moving forward. 


Being unfamiliar with an area makes women unsafe


Being able to connect with other people helps our users to feel safe


People don't reach out to the police when they are uncertain that they are in danger


Users want to reach authentic crime information of a neighborhood


Users want to know about the different qualities of an area which makes them feel safe (ex. illumination)


People prefer to be closer to the areas that are more crowded

04. Understanding Users


Understanding Users

By creating user personasempathy mapsjourney maps, and task analysis we aimed to go deeper into their experiences, thoughts, and emotions by giving our users "a voice".

Surface Book - 1.png
Surface Book - 2.png

05. Ideation


To come up with a wide range of ideas, we engaged in an informed brainstorming session and creativity/feasibility chart keeping our pain points in mind. 


press to zoom

press to zoom

press to zoom

press to zoom

Potential Solutions

After our sessions, we selected three possible ideas then created the low-fi paper prototypes and tested them with our users.


Transportation App & Key Fob

SOLO is an application that comes with a corresponding key fob that has two modes: SOLO and DUO. By choosing SOLO, users can find a bicycle, scooter, or car that they can unlock and drive. By choosing DUO, users can call a rideshare service to pick them up.



Familiarity with other apps.


The phone is distracting.

Not for an emergency situation.

Can't select the route and driver.

Copy of SOLO_Transportation App_Wirefram
Copy of SOLO_Transportation App_Wirefram


Walking Buddy

This application shows other people nearby that are walking to a similar destination. Users can join trips and gain a group of walking buddies to accompany them to their destination.



Familiarity with location based apps.


Potential security concerns.

Creating stereotypes.

Not for an emergency situation.

Landing screen.png


Crime Data & Wearable

This application enables users to reach crime data in an area, and according to that, it provides safe routes. Wearable technology guides them without distracting them.



No need to look at the phone.

Familiar interface.

Using existing data.


Potential privacy & security concerns.

New interaction.





An app and a wearable device that suggests safe routes according to crime data.


The wearable device guides users to the desired destination by using voice and haptic interaction.  


Users will also be able to search for information about neighborhoods and join communities in order to receive that community's latest crime and news stories.

06. Iteration 1



For iteration-1, we created digital low-fi wireframes of our final idea and tested with our users by using think-a-loud and cognitive walk-through methods.

For the wearable, a participatory design session was conducted, and in light of the outcomes, the 3D model was created and tested with users.

Wireframe Testing

The idea was tested by using low-fi wireframes. Cognitive walkthrough and think-a-loud methods used.



Knowing crime data and real-time news of neighborhoods.
Onboarding screens were helpful and fun.
Using existing data.


Needed more information about neighborhoods.

Neighborhood information and news in the routes view.

Several wordings should be revised.

Connection with the wearable should be communicated clearly.

Screen Shot 2019-11-02 at 8.25.14 PM.png

Wearable Design & Testing

To test explore the suitable places and forms of the wearable we engaged in participatory design sessions with our users.


Fingers, wrists, and forearms are the best places to wear the device.

Users designed rings, bracelets, and earrings considering the context and features.

Different statutes and alerts of the wearable indicated by small lights rather than a detailed screen.

Participants' designs.

According to the outcomes of the session, we designed the model and 3D printed it to test the device.

We asked our users to perform tasks and we recorded any hesitations and confusions they expressed. At the end of the session, we administered the After Scenario Questionnaire and questions probing users about the comfort, understandability, and learnability of the wearable. Task completion success noted.




Red and yellow lights are easy to understand.

Turning motion is intuitive.


The battery indication is not clear.

Grasping is hard.

Users don't know the consequences of their actions

A tutorial is needed.

07. Iteration 2


For iteration-2, we revised our design according to the feedback received and created digital high-fi wireframes.

For the wearable, various shapes and designs modeled and the final design is selected. Haptic interaction tests conducted to test the feedback method.

Frame 11 (1).png


By selecting the neighborhoods users can reach their real-time news, crime data, and the overall score.

Users can select qualities that are important for them to evaluate the safety of an area (ex. light, maintenance)

Users can import the contacts that they want to reach in an emergency situation.


1. Onboarding 

Crime Data

Various information can be found on each neighborhood page such as overall score, street view, and news.


The majority of the "overall score" is composed of police reports and a minor percentage is provided from our user ratings.


The crime graphs designed to be easy to understand and under each table source is provided to increase the credibility

Overall crime1.png

Real Time News

Users can see different stories of the neighborhoods.


Furthermore, emergency situations and ongoing instances' progress are provided to our users


Safe Routes

Convoy provides multiple safe routes that users can choose from. Each route has different qualities and ratings considering the overall score of an area.


When convoy + is paired with the app it gives haptic feedback to users to navigate safely.


Wearable Device: Convoy +

The safe route can be uploaded to the device.


Convoy + navigate them by using various stimuli; sound, vibration, and lights


 By using the wearable device or the app users can contact the people who are assigned to different alert modes. 

If an emergency situation happens close to their location, users will be informed and the system will re-route the safe route. 


The safe route can be uploaded to the device.


Convoy + navigate them by using various stimuli; sound, vibration, and lights


 By using the wearable device or the app users can contact the people who are assigned to different alert modes. 

If an emergency situation happens close to their location, users will be informed and the system will re-route the safe route. 


Alarm state



Indicated by the red light.

Location of the users shared with red contacts and the police.



Alarm state


0 degree State


90 degree State


180 degree State

The safe route can be uploaded to the device.


Convoy + navigate them by using various stimuli; sound, vibration, and lights


 By using the wearable device or the app users can contact the people who are assigned to different alert modes. 

If an emergency situation happens close to their location, users will be informed and the system will re-route the safe route. 

Indicated by the green light.

The wearable is ready.

Alerts are not activated. 

Indicated by the yellow light.

Location of the users shared with yellow contacts

 Red Alert


 Yellow Alert

08. User Study

User Study 

In the last step, to evaluate if the outcomes match with our aims of designing the product, we have conducted expert-based testing and user-based testing using interactive prototypes. 


Haptic Interaction Testing

We conducted three tests to test our haptic feedback method. We used Arduino and vibration motors to simulate the vibration that the device will provide. We had several restrictions: vibration couldn't be implemented to our due to the size restrictions. Consequently, vibration tests couldn't be conducted in the context. There is a great possibility that the findings mentioned below may change in real-life settings. 



indication of turning right, two subsequent vibrations, and for turning left, one vibration was used. These indications successfully identified by users.

A 2-second delay between vibrations was required.


Five subsequent vibrations used as an indication of an emergency. This was confusing for our users.


Expert-Based Testing

Prior to user-based testing, we have conducted a Design Critique with an expert and asked him to analyze our system and give feedback on how much the current design meets our objectives. Objectives are provided. 



UI is simple and clear.

The wearable is nice for navigating without looking at the phone.

High-level information about neighborhoods is very helpful.

Alerts are very clear to connect with others but in an emergency.


Colors and fonts don't reflect the seriousness

Haptic implementation to phones can achieve the same intention as wearable.

Detailed information about neighborhoods and the source should be provided

Users shouldn't need to use the phone app in an emergency.



User-Based Testing


4 participants

Frame 3.png

~ 1 hour


20 tasks

We asked participants to complete benchmark tasks. During their performance, we measured: Task success(Yes/No), task completion time, and the number of clicks


The app and object combo is interesting.

Like that sign-up is not asking for too much information.

The aesthetic is very clean.

 Different ways of representing data.

Personalization of the information.

Easy to use.


Users found onboarding screens too long to follow and remember.

Images and short descriptions of gestures of convoy+ in the onboarding process were not enough for our users to remember them.

Turning motion was very intuitive but too easy,

Scrolling took a lot of time. Users ignored the information at the bottom of the page.

The menu bar and back button appearing on the same page were confusing.

Stars represent the safety level but it doesn't demonstrate very precise data.

Need a tutorial video.



noun_click_2397272 1.png

Avg. No.Clicks


Frame 3.png

Avg. Sec. per Task


SUS Score



09. Design Improvement


After the user study design of selected screens are improved.


Typography changed to provide a more clean aesthetic.

Splash screen added to communicate the state of the system.

The size of the buttons and checkboxes increased for easier interaction. 


News Page

News card size increased and tagging system proposed. In this way, users will see the instances close to them.

The size of selections in the neighborhood listing increased. 

Tab design improved and switched to a timeline.

News Details

Tabs added to make the navigation easier between summary, timeline, and photos. In this way, information at the bottom of the screen will not be missed. (1).gif

Neighborhood & Crime Report

The neighborhood detail page provides to provide more clear information to users. Star rating of qualities switched to a bar and supported with colors and texts.

"Crime report" CTA switched to a more visible view. 

Crime report page supported with sticky crime key to increase the understandability.

Comparison of crime between week, month, and year provided.

The color scheme changed to increase the differentiation between crimes. (3).gif

Lessons Learned

10. Lessons Learned

Improvement never stops

After a year later, I did an evaluation of Convoy and realized a space for improvement. I revised the UX and UI of the design accordingly.

"Close enough" contexts are great for getting insights.

Because of the technical limitations, we had to come up with alternative methods to test our design. Even though tests didn't take place in the real context, they were good enough to get important insights.

Frame 4 (2).png