There's a road in Greenville SC that is very treacherous for car accidents. However, it's particularly difficult to know which parts are worse than others. The road is over 20 miles long filled with shops and malls making accidents extremely common. Knowing what areas are bad in this road as well as other roads around you would help you significantly judge your speed and carefulness.

What it does

The system uses big data processing in order to quantify big data and process that based on your current location. By loading the area around you, we can quantify the amount of car accidents that happen within a specified distance to find how dangerous it is in order to better prepare the driver. In the event of an accident the system will notice you were in an accident and send a text message out to your close friends and family telling them about your location.

How we built it

We are using Google's BigQuery in order to process the data surrounding the location of the Dragon board device. We are using a motion detector that is attached to the dragon board in order to detect accidents in the vehicle.

Challenges we ran into

We had lots of issues using the Dragonboard. There is lots of incompatibilies between the software as well as the motion detectors. The Dragonboard's capabilities are similar to that of a Raspberry Pi but are slightly different. Not enough people have experience with the system in order to provide a good community surrounding the device for troubleshooting issues like this.

Accomplishments that we're proud of

We were able to get the motion sensor working and the GPS working on the Dragonboard.

What we learned

We learned how the GPS works as well as the motion sensor that is separate from the device.

What's next for GPUS

We are thinking this could go on bikes, cars, or any type of transportation that's relatively shockproof unless under stressful conditions. In some cases and situations, it could be put inside objects or even humans to tell if shock is occurring.

Built With

Share this project: