MOBILE APP FOR PUBLIC TRANSIT COMMUNITY INTERVENTION

Commuto

The Project

In 2017, I started a project during my UX course at HackerYou to solve harassment issues and discourtesy in the public transit system in Toronto, the TTC. After several prototypes and extensive research, I learned that many cities tried to solve this by making the same mistakes – creating reporting apps that just made relations between riders even worse. I learned that most harassment issues were solved by commuters inside the same ride and, thus, understood that the design solution was to encourage the shift of a negative mindset by enabling people to help each other instead. My initial prototype was selected as one of the course's top projects and was presented at HackerYou's Demo Night.

PROBLEM INSIGHTS

Discovery Phase

At HackerYou, I was challenged to create an app for public transit. I began my research by identifying the biggest pain points for TTC users and discovered driver discourtesy was one of the top complaints. After posting some discussions on Reddit, many commuters defended drivers' attitudes as they are faced with daily challenges, such as dealing with drunks, drug addicts and scammers, all while attempting to pay attention to Toronto’s hectic traffic to safely deliver passengers to their destinations. On the passenger's side, many people experience daily stress from work and personal lives, and while experiencing a packed commute, these negative emotions often surface.

GUERRILLA STREET RESEARCH

As part of my research, I went to a streetcar stop near the intersection of Spadina and Queen Street, a busy area, to ask some questions in order to understand more about the rude behavior and harassment problems caused by TTC drivers. The initial reaction of people I approached was hostile in at least 40% of cases. This confirmed some of the insights I uncovered on Reddit, stating that the reason TTC drivers were rude was that passengers were being rude to them first. Every single person I interviewed said they had never been harassed by TTC drivers, and in one case the driver was rude because the person actually was doing something wrong which caused their negative reaction. Another person told me that service really improved in recent years. I started creating some hypothesis:

  • 01

    Rudeness from drivers is usually motivated by passenger infraction or hostile behavior towards them

  • 02

    Harassment initiated by drivers is rare, and most of them are caused by passengers against other passengers or drivers

  • 03

    A negative and stressful environment creates a negative feedback loop

COMPETITIVE RESEARCH

COMPETITIVE RESEARCH

In 2017, the TTC was repurposing a reporting app based on the failed proprietary app Elerts (white label). Used in different cities such as San Francisco under the name Bart Watch, it failed to generate user adoption. This app allows users to send photos and text messages to the transit police but it had low impact on harassment because the police would not arrive in time to fix the issue. Commuters were also using the app in a way that encouraged racial profiling/discrimination to innocent people and submitting false reports.

SWEDISH CAMERA LOTTERY EXPERIMENT

This experiment was conducted in Stockholm, Sweden. It is not an app but very important for my current research as it deals with positive reinforcement. Cameras that could read car plates and calculate speed rates were placed on a street in Stockholm. If the camera caught a particular car traveling below speed limit they would be automatically inserted into a lottery draw. The money was pulled from others who surpassed the limit and got tickets. The draw would occur and a lucky winner would get a cheque. The experiment was a success and resulted in a 22% reduction in average speed.

FRAMING THE PROBLEM

Reporting apps use negative reinforcing conditioning as a primary psychological trigger to decrease aggression. It uses the hypothesis "if we create a reporting app that allows commuters to police each other we can reduce harassment" but that never worked. In terms of addressing the discourtesy issue, rating drivers would likely not be effective either. I also discovered that most cases of harassment were solved by other passengers moving the harassed person away from the aggressor. This discovery led me to read about the new field of positive psychology. I framed the design problem as "How might we improve passenger relations by using positivity and encourage people to help each other?" I lived in 3 different countries and TTC has, so far, been the most comfortable/satisfactory experience I ever had. I realized that what I really needed to impact was the negative mindset, enabling a shift to positive.

VALIDATING ASSUMPTIONS

Research Phase

After documenting the ypothesis and insights found in the discovery phase I came back to my street guerrilla tactic. I printed out some surveys and engaged with people on the street to ask them to fill out these forms. Tip: if you offer financial compensation and say you are a student people are more likely to participate in the survey. I also recruited Reddit users and others students from HackerYou to ensure I had a more diverse population. Here is a summary of the findings:

  • 01

    Respondents had never been harassed by drivers and experienced very few situations in which the driver was rude

  • 02

    Respondents feel that drivers are the victims by dealing with problematic passengers and high-stress situations

  • 03

    Rude behavior usually occurs as a trigger of a passenger initiating aggression towards the driver or another passenger

  • 04

    Most people do not report a harassment problem to the TTC because they either never experienced it or think there will be no investigation (TTC doesn't care)

  • 05

    Respondents liked the idea of using positive reinforcement to reward good behavior for drivers and passengers. Note: this idea seems to be the strongest in the app and needs to be explored further.

  • 06

    Harassments originate from marginalized populations, such as people with mental illness, beggars and drug or alcohol addicts.

User Research

  • Amy Song

    Busy female in her mid-30s who works in PR and commutes daily via TTC. Agreeable personality and cares about others, but avoids conflict. Her goal is to remain safe and arrive at work on time. Feels that the TTC is too packed which increases her stress levels.

  • James Georgiou

    Assertive stressed out male in his on who drives a streetcar. Experiences constant discourtesy and lack of cooperation from passengers. Often deals with intoxicated passengers but needs to focus on traffic threats. His goal is to keep to his tight schedule and avoid road accidents.

Typical TTC Commuter is female, between 25 to 55 years old and high income

ESTABLISHING GOALS

Design Phase

I started doing a walkthrough of the app by doing phone interviews with Reddit users. Feedback indicated that the solution was interesting but nobody was going to use an app just to help others. Despite having supposedly validated my hypothesis, the app was just not good enough to develop user adoption. I needed more features that were not distracting but useful. Every single person I interviewed flagged privacy and stalkers as a huge concern having their names/photos identifiable. I also needed a more polished design. For that reason I abandoned material design guidelines and started developing my own style, then new design goals:

  • 01

    Locate yourself within a transit line

  • 02

    List drivers and passengers located in the same car as yourself

  • 03

    Send a request for help alert notification to other passengers in the same car when needed

  • 04

    Recognize and reward “good behavior” of specific passengers and TTC drivers within the same car using the same app

  • 05

    Reinforce a positive environment for all TTC commuters within the same car

  • 06

    The system does not report to the police but rather focuses on community intervention

FIRST PROTOTYPE

TEST OUTCOME AND NEW GOALS

Iteration Phase

I started doing a walkthrough of the app by doing phone interviews with Reddit users. Feedback indicated that the solution was interesting but nobody was going to use an app just to help others. Despite having supposedly validated my hypothesis, the app was just not good enough to develop user adoption. To be habit forming I needed more features that were not distracting but also useful.

  • 04

    Avatar generator instead of real photos (increases privacy)

  • 05

    Nickname randomizer (changes username after getting into each car, to improve privacy and prevents users from being tracked by offenders)

  • 06

    AR View - Enable users to track where passengers requesting help are located in real time (since users cannot be identified) using an Augmented Reality beacon

  • 01

    Find a streetcar/bus stop (replaces Rocket Man app)

  • 02

    Autodetect destination based on common trajectories (Machine Learning)

  • 03

    Auto-suggestion feature to improve communication speed and reduce user input

Deployment and Impact

The Outcome

My first project was awarded one of the best cases of the advanced UX course at Hacker You 2017 Cohort. I was called to demo the first iteration on demo night and talk about how I solved a design challenged focusing on behavior and understanding of a problem. My second iteration prototype is still being polished to be finally tested one last round with real people on the streets.