EmerSAFE

Inspiration

UNC Shooting - friends and family were endangered

Our personal safety came to question (<10 miles) Our teammate’s personal experience in a school shooting - zero communication ⇒ more panic
EmerSafe would be a way to calm the panic & create a central form of communication system with police

What it does

Admin’s Side - w/ student’s permission, collects their individual locations in an emergency situation and presents it to law enforcement on a map

  - See where groups of people are located

Client’s Side - every student will be shown a map of the building they’re in, nearest exists, rooms/etc Students can report injuries

  - Will be presented on admin’s side in a different color on the map

  - w/ AI, the location of the shooter can be configured given where the injuries are being reported

Students can also send possible descriptions of the shooter (ex. Gray shirt, pink shorts)

  - Similar reports of the shooter can be pushed out to the admin and client side of app 

  - Students can send pictures/videos of shooter to police

Future Additions / Current draftings

  - Use neural networks and ML to track concerning movements (ex. Entering every room, near injured students) and actions (ex. Turned off phones or location services left at home) to eliminate who the shooter can be, and gain insight into who/where the shooter is

How we built it

  - Utilized xampp and phpmyadmin as a database management system to aid communication between client and admin side of website -- API's

  - Designed UI and UX using Figma for admin and client side

  - MongoDB for potential AI (NN) and data implementations

Challenges we ran into

Geolocating mobile devices that constantly updates itself requires as API that you have to purchase

  - So, in our prototype, we made a photo model of our building (Southgate) to show an example of how our app would work from the client’s end (due to payment issues)

Collecting the locations of students with their permission in an emergency situation would be somewhat an invasion of privacy

  - So, we made the location sharing optional and if collected, only available to law enforcement —- by the Patriot Act

Students sending descriptions of the shooter will require a lot of training for a model to configure which descriptions are real, common, etc.

  - So, we thought about making the options to describe the shooter a drop down menu (ex. Shirt, shorts, pants) (pink, green, blue)

  - If students tend to choose common descriptions (ex. Blue pants), we can reasonably argue that it

wasn’t a fake sent by the shooter to derail the police ⇒ thus push the description out to admin and client side

Accomplishments that we're proud of

  - We are proud of how we were able to connect the mobile app with the desktop/web side using Sampp and phpMyAdmin.

  - We are also proud of how we delegated the different tasks we had to do and worked together on debugging given the time constraints of the hackathon as beginners.

What we learned

  - We learned php from scratch to configure the database with the admin and client side of the app

  - How to compromise when conflicting thoughts are put on the table

         - When one team member wanted to create a central form of communication between all

students as a safetynet for them in an emergency situation, another member brought up how it would lead to spamming, slow app/internet, etc.

         - So we compromised on students being able to report injuries/emergencies and descriptions of the shooter to the police and for them to determine which are real/important to push as announcements to the rest of the students on the client side using AI

What's next for EmerSAFE

  - Purchase a geolocating API that will send live updates to students’ locations in an emergency situation to the admin’s side

  - Create an algorithm using ML to show students and the law enforcement the general location of where the shooter can be based on suspicious movement (ex. Entering and leaving different rooms) and near injured students