The Need for Safety
by Sneha C, Daniel Haim, Jessie Lyu, and Ananya Nandy
This article discovers the need for safety of student personas who are concerned with its concept in and around the UC Berkeley Campus. It intends to discover their Needs & depict a problem statement based on them.
Collaboratively written for the Introduction to HCI class taught at UC Berkeley.
So why discover needs for safety?
We decided to choose safety as our concerned topic because most students don’t feel safe beyond a particular time at night (especially) walking back home or out of campus to the bus. Being on a beautiful and expansive campus as this does come with some cons particularly if you’re not a native of the area. We thought out prime ‘safety’ concerns to ask our users to uncover if they feel the same way and learn what they think about it and whether, down the line, we could build something that could help them feel safer.
Part 1: Needfinding
To begin our discovery on ‘safety enroute home’, we conducted interviews with five students from different backgrounds, particularly those concerned with their safety.
For each user, we interviewed we drew out an empathy map which helped us better define what they said, thought, felt and actually did. This helped us narrow down goals and eventually conclude on our problem statement.
Here’s a link to our detailed observation notes, in case you’d like to take a look at how we interviewed these participants.
User 1: 24-year-old Male
Time & Place: On-campus at 9.15 PM
Our first user was a 24-year-old, 1st-year Graduate student pursuing his masters at UC Berkeley. He’s relatively new to the city (been here for less than 3 months). Lived in Chennai, India for 4 years before shifting here. He currently lives in the Southside of Berkeley (10 min walk) and normally commutes to & fro by foot.
He believed where he lived earlier(In India) was much safer in comparison to Berkeley. “In Berkeley, you need to think twice before stepping after 11 especially in the alleys where lighting is not good”. On-campus, he felt safest at the building he spends the most time in. According to him, the level of danger in the night is much more because of how serious the situation can get (eg. mugging turning into a shooting) which is unpredictable and that’s why people would be apprehensive to help one other. I learned he preferred company while walking back and innately looked out for the person he was walking with as well. He would like to feel safer walking back if he knew which that way was.
Needs:
- Needs to know a ‘safe path’ to get home
- Needs to know where to find well-lit spaces around campus so that he can access them, even at night
- Needs to know he can step out without feeling concerned about being harmed in some way
User 2: 30-year-old Female
Time & Place: On-campus at 6.45 PM
Our second user was a 30-year-old, 1st-year Graduate student pursuing her masters at UC Berkeley. She’s relatively new to the city (been here for about 4 months). She lived in Korea before this. She currently lives in South Berkeley and commutes to and fro by the bus (10–15 min bus ride to campus.)
This user always preferred getting back home early. According to her, she felt “a little safer on the bus, I don’t want to walk back and see homeless people walking around, that makes me feel really scared”. Late for her meant anything beyond 7 PM. In fact, she sometimes even preferred taking a cab over the bus safer as the transport was private. Her feelings mostly indicated how anxious and paranoid she always felt, being uncomfortable and feeling unsafe with the number of homeless people around. She says “I have a bias that they might cause an issue, makes me “anxious” but they don’t really do anything. But I still have fear.”. For her safety is more a “life-threatening issue in Berkeley”
She considered her previous location (before Berkeley) safer too and felt more confidence than here. To protect herself she preferred getting herself a whistle or pepper spray as they were ‘handy’ and easy to access in case needed.
Needs:
- Needs to know when the right time to leave school is before it can be too late for her
- Needs to know where the closest crowded bus stop is and how to get there
- Needs to have a way to keep herself ‘protected’ to feel more confident in getting back home in the dark
User 3: 24-year-old Male
Time & Place: Over the phone at 6 PM
The user is an alumni of UC Berkeley who is currently in graduate school on the East Coast.
He lived on the south side of Berkeley during undergrad, about 15 minutes walking distance from campus. He lived with multiple roommates that were in the same major as him.
The interviewee said that he generally walked everywhere, especially since he lived close to campus. As a student, he did not have to go too many places that were far enough from campus that he could not walk. In addition, whenever he would have to go far, he was able to get rides from friends who had cars. Although he left campus at around 11 PM everyday, he can’t remember times when he ever had to walk alone. For this reason, he never really thought too much about safety while he was a student at Berkeley.
One specific incident he recalled that made him feel unsafe was when he was walking back with a friend and saw a couple get mugged right across the street. He remembers immediately running because he and his friend did not know what to do and were worried that they were next. He later felt bad for not being able to help the couple.
Needs:
- Needs to walk in a group when leaving campus late
- Needs to know what to do when in an unsafe situation
- Needs to be in a private vehicle if traveling a far distance at night
Says: He says that he only rarely felt unsafe because he was surrounded by other people
Thinks: He thinks that having access to cars decreases the worry about safety.
Does: He always travels in groups of people when going to and from school.
Feels: He feels unsafe immediately after witnessing an unsafe event, but safe otherwise.
User 4: 22-year-old Female
Time & Place: Berkeley, Northside at 8 PM
The interviewee is a student in the MEng program at Berkeley. She moved here from Indiana where everyone would drive everywhere.
At Berkeley, she uses a combination of biking and taking the bus to get everywhere. Because her schedule is abnormal, she normally commutes to and from school by herself. Specifically on Thursdays, when she has class until 9 PM, she is uncomfortable going home. One particular spot where she mentions being uncomfortable is the stretch of Virginia Avenue between Shattuck and MLK. For this reason, she never walks home if she is out after it is dark — she will either bike or take an Uber.
The interviewee feels that biking is safer because you can get home faster. Time is important to her and she mentions that one of the reasons she goes home alone most of the time is that she does not want to wait for others or have to coordinate schedules.
A specific incident where she felt unsafe was actually not at night but during the day when someone at a bus stop came up to her and stared at her. She says she does not know why she was targeted even though other people were there.
Needs:
- Needs to feel safer at times in broad daylight when it seems like you should feel safe
- Needs to be able to walk with people only if it fits her schedule
- Needs to know people she is walking with during a commute
Says: She says that she would not feel any safer with Bear Walk because it is a random person and potentially if it’s another girl, it wouldn’t help.
Thinks: She thinks there are specific areas that she considers unsafe and tries to avoid those.
Does: She does not walk if it is dark outside and there is no good lighting.
Feels: She sometimes feels unsafe in broad daylight if someone targets her to talk to her.
User 5: 29-year-old female
Time & Place: Over the phone at 9:20 PM
The interviewee is a Berkeley alumni. She stayed in Berkeley for two years from 2017–2019 when pursuing her Master’s degree. She currently works for Google as a UX researcher.
She lived in south Berkeley for one year and moved to north Berkeley the second year. The first year she lived alone in a shared dorm and the second year she lived with her husband and her mother-in-law in an apartment.
She commuted with AC Transit bus everyday from/to campus in her second year (the first year she usually walked to school). Overall, she felt the bus riding experience was not bad. She normally went to campus at 8:30 AM and came back home before 6:30 PM. When asked about safety she said: “I didn’t experience any unsafe issues on the bus route 52.” When asked about why take bus instead of driving she responded: “Because bus was free and the stops were right in front of my apartment and I-School while driving required parking fee and more walk (the parking lot is off-campus). ” However, she also mentioned that the northbound bus was safer than the southbound bus, and she never took the bus in south Berkeley.
When asked about pain points, the interviewee talked about the uncomfortableness on the bus and the long commuting time. However, she said she solved the issue by preparing something to do on the bus. She also “wished the seats will be better designed so people feel more comfortable on the bus”. She also complained about too many people on the bus during peak hours and the hardness to find a seat.
Needs:
- Need to take the bus in convenient and safe stations
- Need to find something to do to kill bus riding time
- Need to be able to find a seat during peak time
Problem Statement
How might we help people feel safer when they are commuting home?
We considered answering this in a few ways.
Solution 1: Most likely to work
From our user interviews, we gathered that individuals felt ‘unsafe’ when they sensed either or all of the following, primarily speaking:
- ‘Darkness’ or the lack of light on a street
- ‘Empty roads’ or the lack of other pedestrians/other people and vehicles
- No symbol of security or the lack of authoritative or police enforcement in and around
With these conclusions, users would decide whether they would want to take a risk and follow a path they still felt unsafe with or not. We also realized, there was no way for them to know beforehand if there was some sort of danger nearby. Keeping these considerations in mind we made our ‘Most likely to work solution’: “Home Safe”.
We started out flaring from our ‘How might we’ (a reference to double diamond technique), low fidelity wireframing to ensure users could see where other users like them were, had a way to guide/navigate themselves based on their current location and new how much time the ‘indicated safe path’ would take them to get home. We also considered providing an option for users to set off an alarm in case they felt they were in danger (interaction idea: set off the alarm by double-clicking the power button quick so they can access it even if the app wasn’t in the foreground & the phone’s screen was off!)
Narrowing down to the High Fidelity Prototype (Using Figma), we focused on the experience from launch to achieving the final goal of a user as a MVP(Minimal Viable Product) Users to just have to enter their phone number(to allow us to send a tracking link to a trusted person of their choice) & enable location(so we can pick up their surroundings and detect ‘safe paths home’ to begin. Once in, they could enter where they would like to go and based on what we classify as safe(the conditions above) we would indicate paths for them to get Home Safe.
Solution 2: Most rational
The second solution, Stay Informed, tries to keep students informed about unsafe situations around the campus so they can avoid those areas in commute. We realized it will be helpful to create an online forum so people can share their “unsafe” stories. The app will generate a map with all the incidents mapped and indicate safe and unsafe paths based on the number of incidents.
Low-fidelity prototype
On the initial ideation, we focused on creating an online forum so people post their experiences of “unsafe”, such as darkness, meeting strange people, getting robbed, etc. The users could tag their incidents for categorization. The app will also generate a map for users to view where most of the incidents happened.
High-fidelity prototype
The high-fidelity prototype is similar to the paper sketches. However, we decide to limit the login mechanism to only CAL students so users need to log in with their CalNet ID. This way we can keep this virtual community clean and safe.
Some selected screens
We believe by creating a crowd-sourcing platform, users will be better informed regarding dangerous situations nearby. Additionally, this application could address the dynamics of “unsafe” places as people keep updating the incidents.
Link to Figma prototype.
Link to interactive prototype.
Solution 3: Most Likely to Delight
From the user interviews, we gathered that users felt safer when they were walking with another person, but that they would not want to sacrifice convenience or walk with a random person. While some mentioned that they heard of BearWalk being used frequently, others mentioned that they weren’t really comfortable with meeting a random person to walk with.
For this reason, we came up with the idea of BearBud. Instead of a police-sponsored safety walker, the idea behind this was to create a social way to stay safe. Specifically, we wanted to build from the idea of a social network and use it to match the user to people with similar schedules who were also heading in the same direction. This way, people can develop friendships as well as increasing safety by having someone to walk home with.
The BearBud app would be limited to people who go to UC Berkeley in order to make the network more constrained. For this reason, after login, the user would have to authenticate using the CalNet system.
The users would then input information about themselves, their schedules, and their home locations. The app would store the location information as a map radius instead of an exact address in order to maintain privacy. Then, the users would be provided with a list of matches according to schedule and location. They could then message the their matches to set up the walk home!
The following screens show the very basic functionality of matching users with people with their schedule so that they can coordinate walking home together.
Acknowledging that schedules may change, especially for a student, we also wanted to incorporate situations where the user might want to look for someone to walk home with them on the spur of the moment. For this reason, we incorporated an instantaneous match page. This would be useful for one-off times where the user is going home at an unusual time. By maintaining the user data, the app could check if someone the user has walked with before is also leaving at the same time. In addition to messaging to initiate the walk, we included a simple request as well, for those who are more shy. This workflow is shown below.
Users 3 and 4 could use this in slightly different ways. User 3 already has a group of people to walk home with. By joining the BearBud network, the user could include someone else who is walking the same way, who would now benefit from having a group to walk with. User 4 does not walk with people for convenience reasons as well as not wanted to walk with a random person. With BearBud, the user can find people who match her leaving times exactly and consistently walk with them, so she can build a relationship of trust and maybe even friendship with them.
The link for the full interactive prototype is here.
Solution 4: Blue (Dark Horse)
Blue (named after the blue emergency lights) is an app that allows users to see and interact with other pedestrian commuters that walk to and from campus for the sake of creating a safer commute. The users log in to the app using their school ID, which provides identity validation. After the user validates their identity, they enter a game like voice chat that allows all users on the app to communicate. The users can see navigation directions and other users’ locations on a real works AR map (just like a multiplay video game). If the app identifies two users that share a similar route it will match them so they can walk together for some portion of their route.
The app also provides a way to call for help, this can be done in multiple ways including, pre-recording an emergency interaction pattern with the phone (like sequences of button clicks, or prerecorded swipe stroke) or saying a keyword the triggers a calls for help. The call for help can be configured in two ways, the nontrivial way is requesting other users who are within a certain radius to run and help the user that sent an SOS signal the other configuration is the typical connection to policy or other authorities.
Usability Testing
Recruitment Strategy: Ask students who are concerned with their safety if they would be willing to give feedback for a solution to help them feel safer and make their surroundings safer for themselves and others.
Goals:
- Determine what information should be provided for students who are concerned about safety on campus.
- Determine the best way to present the information.
- Determine whether to incorporate a social aspect to the solution.
- Determine whether this app helps them feel safer by showing them a ‘safe path,’ by allowing their location to be shared while they head home, or by calling/creating an alarm for help when they need it
- Determine whether they would rely on either of these solutions to achieve a sense of safety
Questions:
- What is the most important piece of information for you to know when you’re getting back home at night? [tested by Jessie]
- Would the user be more likely to use a social or individual app when heading home? [tested by Ananya]
- How safe does the user feel using this app to help them get back home vs. this other app & why? [tested by Sneha]
The general usability study plan and guides can be found here.
Test 1
What is the most important piece of information for you to know when you’re getting back home at night?
Participant: 2nd-year I-School student
Prototypes: Informed! and HomeSafe
Task 1: Exploration of Informed!
- What do you think the app is for?
- How do you find out if it’s currently safe for you to walk back home?
- What would you do if you find an incident that just happened on your way home?
- How do you plan the best route for getting back home?
Task 2: Exploration of HomeSafe
- What do you think the app is for?
- How do you find the safest route to get back home?
Task 3: Scenario
You have a late discussion today until 8 PM. When you finished the meeting it’s already very dark outside. You live in Telegraph & Durant. You are worried about your safety getting back home. You want to know if it is ok for you to walk home alone.
- What information do you think you need to get before starting heading home?
- How can you get that information?
- Once you obtained the information, will you feel safer walking back home?
Results:
For both applications, the interviewee was able to figure out the functionalities of them. All the tasks were conducted smoothly and much insightful feedback was collected.
For task 1, the interviewee figured out the app was a forum to track incidents happened around the campus. However, it took the interviewee some time to find the “incident visualization” feature of the app. The interviewee thought this feature could be more visible because she felt it was useful for her to walk back home. When asked about how she figured out if it is safe for her to walk back home and how she planned the route, she said she will use the search and tag. But she also felt the tags were very arbitrary. She also said she will use the map to see where was unsafe but she mentioned what if all the paths to her apartment were marked unsafe. She said she will probably feel even more unsafe if she sees that many incidents happened near her apartment. She also mentioned the dynamic of the app as she said some people might update the incidents with quite some delay and the map was actually outdated at the moment she tried to use it.
For task 2, the interviewee felt the purpose of the app was unclear in the first place. The interviewee thought the app was like another Google map until she entered the navigation page. She also mentioned that the text is a little small for her. She had a little trouble figuring our the representation of the brightness of each road. Since task 2 was conducted after task 1, the interviewee inevitably started to compare the two prototypes. She asked why HomeSafe used phone numbers as credentials and she mentioned that using Berkeley credential will be safer as it will restrict access to only Cal students. She also mentioned that the map functionality on both apps was similar and she said it will be better to combine them together. She also said the text color was too light and she had difficulties reading it. Another thing she mentioned was “why doesn’t it(the app) show safety info on my path but on some other roads?”
For the final task, the interviewee chose to use both apps to find the best route. She mentioned again it will be way easier if the two apps could be combined. She also mentioned that Informed! was similar to Wildfire, another community app for students to receive updates about safety incidents. She had trouble searching for posts on the first app because she didn’t find the appropriate tag. And she said it will be better if the HomeSafe could automatically plan the best route for her.
Reflection:
From the usability test, we found that the most critical information for this user was both if there are incidents happened just now and safety amenities. Also the feeling of “safe” is important for the interviewee.
Test 2
Participants: 1, 22-year-old female
Recruitment: Follow-up from needfinding interview
Non-participants involved: None
Setting: 8PM, North Berkeley
Question: Would the user be more likely to use a social or individual app when heading home?
Prototypes: BearBud and HomeSafe
Tasks:
- Exploration of BearBud
- Exploration of HomeSafe
- Scenarios
You have been working hard on your project and now it is already 9PM. There are still some people out and about on campus as they are returning from a football game but most of them are going to the Southside while you guys need to go to the Northside. Walk me through what you would do. You can choose to use any functions in either of the prototypes you explored.
Let’s imagine your friend John. John has a pretty long walk home from Etcheverry Hall to his apartment on the outskirts of Berkeley, close to Oakland. He is a very friendly person, but is new to Berkeley and doesn’t know many people yet. He also isn’t completely comfortable with where everything is around Berkeley yet. However, he recently got very scared during an incident near People’s Park where he thinks someone was following him. Now, he’s looking for something that will help him feel safer. Would you recommend any of these apps to him? If so, which one? Why?
Results:
In the usability test, I had the user use the two prototypes to determine which functions were the most important regarding safety. I used BearBud as a representation of a solution that was more “social” and HomeSafe as a representation of a solution that was more “individual.”
The user’s interactions with BearBud were interesting because she used it differently than expected. For instance, when going home late during an unscheduled time, instead of using the function for finding random people who were also walking home, she just added the new leaving time as an event to regenerate the matches. This indicated that she would be more comfortable with the people she was “matched” with rather than a random person, even though in reality, all of them are random people. The user also stated that “[she would] check it everytime [she] walk[ed] home but not late at night because [she] wouldn’t walk home that late even with a friend. Maybe it was a big group but not really for 2 or 3 people.” For that reason, when she saw that the messaging was between pairs of people, she said she wished there was a way to coordinate the same thing with a larger group.
With HomeSafe, she immediately thought it was a good idea because she had never lived in Berkeley before so didn’t know it wasn’t safe. For example, she wouldn’t have known that People’s Park was bad and would not have known to look for that information. Therefore, if other people were using app she would use it too. She didn’t think most of the information was that useful except for the areas to avoid, which she would look at at the beginning of the trip. This was because sometimes, you have to go down streets that don’t have lights and she would not walk extra just to avoid that. She did not seriously focus on the information about other users being out and about on certain streets, but said that she was unlikely to use the app for the duration of the trip.
When choosing between the two during the scenarios, she emphasized that “if you have more people, it feels more secure than if you just know the path of least danger.” While she mentioned that she liked that BearBud has other aspects than just safety, this was secondary to the theme of safety in numbers. Therefore, while the user did place importance on the “social” aspect over the “individual,” it is possible that an emphasis on the activity of other users (and having them be part of an “in group”) on streets in HomeSafe would also satisfy the user’s needs.
Detailed results can be found here.
Reflection:
Based on the usability test, it seemed like the best solution would be something that can be used at only the start/endpoints instead of during the walk. Having safety in numbers was important but having the information be passive (they are not part of your group) was not as effective.
Test 3
Participants: 1, 25-year-old male; Non-participant: 1 interviewer
Recruitment: Follow-up from the need-finding interview
Setting: 6 PM, On campus
Question: “How safe does the user feel using this app to help them get back home Vs. this other app & why?”
Prototypes: BearBud and HomeSafe
Tasks:
- Exploration of BearBud
- Exploration of HomeSafe
Overview:
- The usability test was testing the question: “How safe does the user feel using this app to help them get back home Vs. this other app & why?”
- On this basis, we showed the user two interactive prototypes which required engagement from the user but also provided information from the user's surroundings to help them feel safer
- The user was interested in participating without an incentive, but with the intention of knowing that there was a chance to feel safer when heading back late at night as they have faced this situation themselves.
- The user was very interactive with the prototypes & explored the app in its entirety, asking very relevant questions
- Usability issues observed are listed below
- Findings & conclusions of the outcome of the test are also summarized from the interaction
Prototype 1: BearBud Task
Background of the user:
Users head home late at least twice a week. His definition of late is past 9.30 PM. He doesn’t have a preference for walking home with someone or not but says it's a ‘50–50’ in a week between how many times he walks alone & walks in a group(up to a certain point). They claim to feel unsafe primarily after 10/11 PM because the streets are deserted and the chance of miscreants is higher. He seems to believe that there’s also some amount of ‘luck’ involved in his encounter with such a situation. His first instinct in an unsafe situation is to call his roommates & head to a place where there were more people around so they could help.
Login flow & Initial Schedule & Location Screen:
- Asked whether he had to log in using email — on tapping found out he can just proceed, seemed like his first time with a rough prototype
- I was unclear with the flow — but got into the home page. Was confused because wasn’t sure which schedule was required
- Whether he was supposed to select a schedule of his classes or schedule of his commute
- Time slots provided were not corresponding to ‘commute schedule’ because they were too large in the UI
- Believe ‘commute schedule’ is more relevant
- Was able to navigate to & fro with ease
- Takes an assumption: Commute schedule, because it depends on personal preferences (the time they like to spend on campus irrespective of classes)
- No timing here, no personal profile, which order is it sorted in?
- Makes more sense to coordinate with one person
- Asks for users location — was able to understand
Recommendation:
- ‘Add schedule’ terminology has to change
- Coherence in language & timing is required
- Timing slots could be interactive for users like new age calendar apps allow
- Accessibility to users own profile & home page need to be available because the user doesn’t understand what they have to do
Matched profiles & Chat screen:
- Is slightly confused but not sure how to proceed — clicks on multiple places to ‘discover’ what UI element makes sense and then says ‘it was pretty straight forward’
- Felt the chat screen was very intuitive
- Wasn’t sure how to see the other person’s profile from here and would like to do that before beginning a conversation with the person to have some background context
- Wasn’t sure which order users were shown to him
- Was looking for the timings of the other person’s schedule not just ‘days’ they were available
- Would mainly base their choice of selecting the person on their profile & timings that worked well
- Liked that you build a personal connection with just 1 type of person
Home Screen:
- Discovered this much later by mistake, needs to be more intuitive
- Did not understand the dialog popup
- Would like to access home profile not only from the dialog
Overall Results & Recommendations:
- Include access to other person’s profile
- Indicate their available timings
- Indicate where exactly the person destination will be
- Indicate a ‘Favourite’
- Include other person’s expertise of a destination, if they were going somewhere new.
- Have to ensure that it doesn’t turn out dangerous for women or men in the long term if someone feels uncomfortable
- User said: “
Prototype 2: HomeSafe Task
The user was asked to go through the app and walk through to use it as if it could help them get home safe. The below lists the observations from the study:
Login Flow:
- The user found the flow pretty intuitive and was used to it as he found it familiar to apps he’s previously used
Path Screen:
- The user looked at the screen and absorbed the information, by each indicator and was able to understand was the relation was to its visual elements. Found all the information useful
- Found a lot of indicators which drew his attention, as soon as he looked at the screen
- Route stood out most to him and second the symbols/indicators
- Would have preferred the ‘lit path’ indication on top
- Arrow indication required on bottom screen — tap on the map
- Was able to discover sharing his location & call for help through the app
- The user would use his phone 60% of the time, but check the app only before he began his journey home because once he knew the app he would stick to it.
- Would only rarely use the app after he felt accustomed to space
Overall Results & Recommendations:
- Include easier discovery of path user has to understand
- Allow users how to interact with the map
- Discovery on how the app can be made more helpful is required
- Even though he preferred the use, intuitive nature and feel of the app he would not use it after he was accustomed to the location he was commuting on a daily basis
- Reflection:
Users words: “Utility is subjective of a user inputs. First app can be super helpful, especially late at night. Since other resources of help may be busy. This could be a good proxy as users can connect to one another. People who frequently take a route — they wouldn’t look at the second app after a first time use”
