Janus

Be safe, walk boldly! Janus helps you choose secure routes for your walk from Point A to B by calculating the safety of possible paths.

Comment

Pitch deck: https://pitch.com/v/Janus-9xffye

Inspiration

We were inspired to create Janus because many people don't feel safe walking in NYC, including our friends. As students, we often hear about shootings and robberies happening near our campus. Surprisingly, 7 out of 10 New Yorkers worry about becoming victims of violent crime.

One of our teammates, Hansaem, shared her experience of feeling unsafe, especially as a woman and as an international student unfamiliar with New York. She didn't have a way to know which areas were safe or not.

We think a lot of people in NYC, especially women, can relate to Hansaem's story. That's why we made Janus - named after the Greek god that oversees passageways - to provide a solution to this issue.

What it does

This app offers three possible routes from point A to B: the shortest route, the safest route, and a recommended route in between. Each route has a safety score determined by the number of crimes near the areas the route crosses. With the help of Wolfram Alpha, the weather of the route is also shown.

This app contributes to healthcare by reducing the risk of accidents, injuries, or stressful encounters during walks, ultimately promoting physical and mental well-being.

It can be especially valuable for daily workers who need to commute safely to their workplaces, patients who require safe routes for outdoor activities or rehabilitation, and anyone concerned about their overall health and safety while walking.

How we built it

It calculates the safety of routes by tallying the number of crimes reported near a route, then assigning a safety score inversely proportional to the number of crimes. This approach divides NYC into grid units and creates a base safety score for a route. For each unit the route goes through, if the route goes through the unit and the unit has a crime reported, then the safety score of the route decreases. If the unit doesn’t have a crime reported, then the safety score is unaffected.

Challenges we ran into

The main challenge was finding a strategy to determine safe routes: We initially thought of calculating the safest routes by calculating the route with the largest sum of distances from crimes. However, since we’d need to calculate the distance between tens of thousands of points, the task would be too computationally intensive.

Other challenges:

  • Issues with rest API calls
  • Issues on how to divide the map into grid units
  • Issue displaying google maps API
  • Issue integrating python code to js front-end

Accomplishments that we're proud of

Coming to our solution after hours of thinking and debating! Also:

  • Used Wolfram Alpha to extract current weather conditions for a route
  • Utilized NYC open data to get crime complaint and arrest information
  • Divided NYC map into grid units

What we learned

  • Wolfram Alpha API
  • Google Maps Platform API
  • Data cleaning and aggregation
  • Javascript, Python, and UI/UX

What's next for Janus

Since a user chooses between 3 routes, Janus can learn the user’s preferences for routes by recording the user’s choices as input for an ML algorithm, which can then fine tune the “recommended” route for the user.

Janus can also scrape live crime data through social media and add them to the map, highlighting the details of the report when the user clicks the icon.

We will work to improve the algorithm and launch a mobile app.

+ 4 more
Share this project:

Updates