I was watching a Youtube of a police officer chasing a criminal. They were exchange gunfire whilst driving. Throughout the video that officer was updating the dispatchers on his situation through radio, calling out intersections as he passed them.

Chasing a crimal is a lot like a real time strategy game. The most important advantages are communication and speed. I was amazed that such a dangerous line of work still uses an outdated form of communication.

What it does

Ride-along is a interactive web application that succintly conveys important information to the police officers that use it. Ride-along uses NLP to detect chatter over police radio and plots the important information on the map. Police officers will be able to easily understand the state of their colleagues and will be able to work together better.

How I built it

Ride-along is built using 3 different packages. The client, the radio, and the server. The client is responsible for displaying the events, the locations, and the states of all the officers on the network.

The radio is another webapp that is responsible for converting speech to text, then using our NLP models to detect intents in the police officer's speech.

Finally, the server is responsible for requesting data from external APIs like directions between waypoints.

Challenges I ran into

The biggest challenge was to figureout the best way to skip under API limits. In Ride-along, police officers have the option of navigating to other polices officers. Getting the directions everytime these officer's locations change can use a lot of API calls. Thus why I built the central server to make the calls and store the resulting geometry on the firestore DB

Accomplishments that I'm proud of

Built a full featured webapp over the course of a weekend!

What I learned

Learned how to to use Mapbox and Dialogflow.

What's next for Ride-along

We'll see!

Share this project: