In today's world, we have lots of data that can be used for good caused. We aim to save lives by using Geospatial big-data analysis of data collected during people's drives.

What it does

Our app has two parts- One alerts the authorities of dangerous junctions where many people have sharp changes in speed and has many car accidents in that area. The second part uses a personalized learning algorithms that learns where the driver needs to be alerted before places he is driving in an unsafe manner. This gives a comprehensive solution to risk mitigation of car accidents.

How we built it

We analyzed the data using python. Criteria for dangerous places was when speed dropped to 1/4 of the original speed or less within 4 seconds (original speed >=5= km/h) Two android apps were built to alert the driver- the alert is based on sound to prevent additional distractions. The more advanced app shows the driver its location on the map and alerts him when the he approaches risk areas.

Challenges we ran into

Data was not clean, server was down and the arcGIS API didn't work at some of the times

Accomplishments that we're proud of

High correlation between our criteria and another dataset of car accidents in the UK. It proves that our criteria provides this useful information while also providing data about the driver's behavior during driving.

What we learned

The data available today can be used to help decrease the number of car accidents. Having a larger dataset can lead to even better results. This work is easily scaleable.

What's next for SafetyMatters

Using weather conditions for smarter alerts - requiring lower speed during hard weather. In addition, understanding how this app can run on the car's hardware.

Share this project: