Project is inspired by a real life situation, where one of our teammates was taking an Uber back home and a driver at first seeming like decent person, later turn out to be drunk and our friend had leave the car in middle of nowhere in the midnight.

What it does

It monitors and tracks behavior and patterns of drivers. Giving the employers valuable inside into their driving style nearly in real time. We use sensor data from the phone (like location, camera...) and use machine learning to analyze the output, which is then presented in our custom made analytics portal.

How I built it

We mock the phone with a python desktop application which already does some data processing in order to securely and optimally send the data to our backend service. Data is them processed some more and stored. Our analytics portal receives data from the backend and displays it in a clear and readable manner.

Challenges I ran into

Implementing the backend service with all the data processing would take too much time.

Accomplishments that I'm proud of

Getting the camera based prediction model to work well and building readable analytics portal which presentable enough for the demo

What I learned

We learned how difficult it is work and combine multiple technology stacks together under very little time.

What's next for Hacar

We want to validate our idea with professional drivers and businesses as well as improve upon the prototype, eventually creating a startup out of it.

Built With

Share this project: