Oh Deer

Why: It's hard to say that you don't know anybody who hasn't ever hit a deer while driving around Michigan. In an effort to save lives (human and deer alike), we decided to make an app to help prevent potential accidents.

What: Oh Deer quantifies the danger of hitting a deer throughout your drive. By taking into account previous accident history, current location, trajectory, time of year, and time of day, we have created the perfect app to help drivers stay aware and identify dangerous driving conditions.

How: With the power of open source Michigan car incident data, we can identify dangerous deer locations during your drive. The data is stored on a MySQL database running on IBM BlueMix. The mobile app, developed for Android in Android Studio, connects to the database via a PHP file hosted on the PHP app service on IBM BlueMix. Using POST information provided by the phone, the database will send the most relevant incidents in about a 50 mile radius around the current location of the phone. Oh Deer creates a subset of those incidents which are in a 3 mile radius around the driver, and also extends it 10 miles in the direction which the car is driving. The number of incidents found is combined with information on the current month and the current time of day, to estimate the possibility of a deer related collision. It updates periodically throughout the drive, so that you can see changes in conditions.

Challenges: Getting the phone to connect to the database was particularly frustrating because of different PHP versions between development and deployment environments. We tried many solutions to host the PHP file, including personal laptops and raspberry pis, but settled on BlueMix.

Accomplishments: We made an app that can actually help people. It's not quite streamlined, but the idea is alive and well in our proof of concept.

Learned: A lot more PHP, much more familiarity with making android apps, and that there are a lot of deer accidents!

What's Next: Michigan isn't the only sufferer of deer attacks! This can easily be expanded to included all of North America, simply by expanding the one table in the database. We can move beyond just identifying bad areas, but also providing alternate routes around heavily populated deer areas.

Built With

Submitted to

SpartaHack 2017

Created by

I mostly worked on the mathematics and statics related to the different features in this app.

Zach Taylor
Lordy
I set up the backend (MySQL) server and imported all of the accident data, after parsing out all of the non-deer related accidents. I also created the PHP file and set up the hosting on apache, so that we can interact with the server from the phone. I came up with the idea and pushed the team forward in development!

Brett Durlock
Graduating senior at Michigan State University, studying Computer Science. Will be working at Lockheed Martin after graduation.
I did the mobile app

Tyler Smith