SaFepath

Feel uneasy walking through San Francisco? SFePath.ai helps you find the safest walking routes in real time, using data on crime and local conditions, so you can move through the city with confidence.

Comment

Inspiration

We drew inspiration from a common concern many members of our team have faced walking through the city of San Francisco, safety. Due to the increasing crime rates in the downtown area, many citizens have to carefully traverse the streets of the city, planning out the safest route to take from and to their destination. So, we wanted to build a product which helped users get to their desired location in a safe manner.

What it does

Our application receives the current location of the user, and choose the safest, reasonable path to the target destination, based on previous and current crime incidents.

How we built it

We built the frontend of our application using React.js, and gathered efficient polypaths between our two target destinations through Graphhopper API. From there, we use publicly available crime and active case data to create 'danger points' on our graph, and calculate the most efficient path away from our danger points through graph segmentation.

Challenges we ran into

We ran into issue of initially understanding and employing AWS Services, along with discussing what type of function would be used to calculate the risk associated with each path. The main discussion we had concerned whether to use a pure predictive model based on past incidents or a live model which fetches and uses current data to generate 'danger point' which increases the risk of our path.

Accomplishments that we're proud of

We're proud of creating a front-end, along with creating and applying a modified Gaussian kernel distribution to classify the magnitude of the risk associated with a weighted danger point a certain distance away from our pathway. We experimented and gained a high level of knowledge on multiple AWS service like S3, Lambda, Secrets Manager, CloudWatch, IAM, and API Gateway.

What we learned

We learned to create an intuitive web application, along with with how to work with polypath data. We also learned line segmentation, databases and applied mathematical modelling.

What's next for SFePath.ai

For the future, we hope to expand this project to other high population density areas, such as LA or NYC, as the residents of these cities also face similar challenges facing high levels of criminal activity.

Share this project:

Updates