1 out of every 6 American women have been victims of sexual assault or attempted sexual assault. In 80% of the cases, the victims knew the perpetrator. Moreover, only 5% of reported cases result in the arrest of the perpetrator. In this context, we created pear: an easy to use, simple, anonymous resource application with trauma informed help for survivors and resources to keep people safe in their day-to-day lives.
What it does
pear. has a series of functionalities, including: - Grounding exercises: Functions to calm users down and provide helpful tips in Emergency Situations. This methodology has been used by trauma informed specialists and it ensures that the user can focus on their environment instead of their own thoughts. - Helplines: A easy-to-access list of helpful resources where people can find help and support. Those resources include victim advocacy organizations and emergency text lines. Moreover, we included an one-click automatic messaging to all of the user's emergency contacts -Emergency resources: We included a map (with the option of navigation) of all nearby hospitals. - Navigation to safe locations: in case of an emergency when a user is walking on the street, pear. will provide navigation to the best "safe location" close to you - all with one click. "Safe locations" may include restaurants or hospitals. In order to find the best safe location, pear. takes into account both the proximity to the current location of the user and the safety of the trajectory there, using AI to find the best possible place and route. The navigation includes a map with the highlighted path, as well as an audio giving the directions - Incognito Mode: pear. allows you to go to incognito mode while using the app. At any time, you can click one button and go to a page that mimics a food delivery app, which guarantees the user's safety and privacy - Angela: Fake caller app for the user to safely get away from uncomfortable or potentially problematic situations
How I built it
Design Phase The entire wireframes were designed by our UI/UX expert Sally. All icons were drew by our team.
App Building We used Android Studio and Google Flutter - using the programming language Dart - to make an app that would work in both Android and iOS devices. The program follows a hierarchical code structure, with different page routes having their own classes. We used function calls to utilize app permissions, and various dart packages - such as Url Launcher and Audio Players - to extend the app functionalities.
AI Model To find the directions to the best safe location nearby, we used an AI Model. We used Socrata to get data from the Baltimore City Government Crime Data API. This data includes the coordinates of all crimes with victims that happened in Baltimore in the last year. Moreover, we used Google Places API and Google Directions API to get a list of nearby "safe places" - restaurants and health centers - close to our user's location and the possible trajectories to those places. Using an A* search, we calculated a tree with all possible movements the user could do to go to the possible safe places. Each state's heuristics was given by its proximity of all crime points in the region and the cost of performing an action - that is, going to one state to the other - was given by -1 * distance to the end goal. Then, by maximizing the total cost, we were able to find the most optimal trajectory.
Challenges I ran into
Data It was really hard to find proper data for the AI model - specially the crime data and the city map. Eventually, however, we found good API that gave us really good results. Time We had a lot to do in the app and, at the end of the first day, it seemed like we had little done. However, by parallelizing and prioritizing tasks, we were able to get done with the main things we wanted to accomplish Google Flutter Bugs We were learning Google Flutter as we go, so there were naturally a lot of bugs. By helping each other and googling a lot, everything eventually came to place Sleep.
Accomplishments that I'm proud of
We were really proud of finishing up on time and of our great team dynamics. We believe we have skills that complement each other, which made this app work out. Moreover, we are really proud of making something with such a huge social impact
What I learned
We learned a lot about app development and how to find our way around new technologies, APIs and packages. Moreover, we learned how to effectively work together to make up for the time constraint and sleep deprivation. We also learned a lot about sexual assault and trauma-informed care
What's next for pear.
- Improving the user interface and design of the app
- Adding settings to allow for users to add an emergency contact
- Improving the link between the AI model and the app (and make it run faster)
- Adding a augmented reality support animal as a therapeutic approach
- Adding other functionalities according to the input of the users