Inspiration
Why do people have car accidents? The answer is simple: less experience, not familiar area, unnoticed traffic lights or signs, bad weather, vehicle or bad road conditions. In summary, people lack of information about the certain road segments under certain conditions.
Imagine the situation. You google how to get from A to B by car and you are given a route. Additionally along the route the zones that have a high probability of a car accident are colored red. This means you already have a prior knowledge about the dangerous road segments and you can prepare yourself for it by slowing down, not accelerating and you will pay more attention to the signs and other cars.
What it does
We have built a prototype app (Safetify) that uses a machine learning model that takes traffic, driver, road and weather data into account, predicts how dangerous a particular road segment is and alerts you.
How I built it
The probability of a segment to be a dangeour zone was computed with neuronal networks. The area of the dangeour zone was computed by using GaussianProcessRegressot to smooth out the probabilities and provide a contour. The prototype was build though the API.
Challenges I ran into
We had to figure out how google API works without entering our own credit card information.
Accomplishments that I'm proud of
We planned thoughtfully the whole app idea from the very beginning , coordinated the work and made an amazing teamwork.
What I learned
The usage of machine learning libraries and how to build and use API.
What's next for Safety
Only forward!
Log in or sign up for Devpost to join the conversation.