After hearing this year's Keynote speaker's speech about helping the city of Detroit overcome its obstacles through the pursuit of free and public data, we decided to follow his lead and to address a major concern for locals and visitors alike: safety. Many areas of the city are largely empty, and it's easy to feel dangerously alone and unprotected while walking through them -- especially at night.

What it does

Safety in Numbers uses cameras distributed at intersections around the city. These cameras can recognize human forms, and track how many people are in an area as well as the intervals between their passing. That data is then fed into a live heat map accessible at our website displaying crowd density across the city of Detroit in real time.

How we built it

For the cameras, we attached a Logitech webcam to an Intel Edison computer, and then fed the data into a node.js server, which in turn passed it to a Firebase database. From there, we took the data from each camera and used its location to plot points on a map powered by Google Maps, which used the data to form a heat map.

Challenges we ran into

Of the four of us, three had no prior experience with node.js and server-hosting. To complete our project, we had to learn both from scratch, which took a huge percentage of our time. We also ran into numerous hardware difficulties, ranging from pulling data from the Edison to broken equipment.

Accomplishments that we're proud of

This was one of our biggest projects, and for most of us it was our first foray into using technology for social impact on this kind of scale. Really, we're just happy to have made something to try and help the citizens of detroit.

What we learned

What's next for Safety in Numbers

At the moment, Safety in Numbers is primarily just a reference for people to look at. In the future, we'd like to pursue integration with car GPS modules in order to plan rides through the city and sticking to streets where it's ok to stop at stop signs. We would also like to pursue Facebook messenger bots that could let users know if they are entering a low crowd-density area.

Share this project: