Safety Net

A map of individuals in need of assistance
Example of the first image classifier

Inspiration

With the recent scare of Hurricane Florence, we wanted to create a system to more quickly identify people needing aid in disaster-stricken areas.

What it does

Safety Net is an application designed to run in cooperation with a swarm of drones. Search and Rescue drones are already becoming commonplace, and optimizing their capabilities has the potential to save countless lives. Drones that are deployed to an area affected by natural disasters are often sent to find specific missing persons or to investigate infrastructure failures. Safety Net expands the reconnaissance abilities of the drone swarm by identifying any individuals that are possibly remaining in the region. No prior information is needed about the area - the drones simply sweep the affected zones and use Haar Cascades to spot people in need of rescue on the ground. Safety Net then refines the results using IBM-Watson before transmitting its findings to a base-of-operations in the area. From there, a local server fitted with Google's Maps API drops pins wherever people are found, and the operator can quickly view still images of the area to identify people in need as quickly as possible.

How we built it

We used OpenCV and IBM-Watson as our two visual classifiers. The location data and images are sent to a Node JS server, which communicates with the Google Maps API to display a comprehensive map to the operators.

Challenges we ran into

None of our 4 team members were too familiar with Node JS or front-end development. Much of our time was spent simply exploring the available functionality.

Accomplishments that we're proud of

In line with the previous question, we were very proud that we were out of our comfort zones for much of the time, and yet were able to have a finished project with the intended functionality. Although it is not perfect, Safety Net does its job well and could potentially be developed into life-saving software.

What we learned

As previously mentioned, we learned a lot of new technologies.

What's next for Safety Net

We'll continue to develop the project, cleaning up the interface and adding reach goals until we're satisfied with the relevance and helpfulness of Safety Net.