No-Traff*cked

Inspiration

While commuting back from my college, I loved it when I got a streak of traffic lights green & t eased out my accumulated stress throughout the day. Our Idea helps drivers in a similar way by telling them the optimal speed he/she should drive on to get traffic signals green

What it does

The No-Traff*cked app provides user with the speed that he/she should drive at to avoid traffic lights. In order to help ameliorate user's journey experience & thus resulting in decreasing the mental stress of the person. As per the Economic Being exposed to the daily hassles of traffic can lead to higher chronic stress

How we built it

Real-Time Traffic Light Status We will deploy an edge device connected to existing camera's placed at traffic lights, which will have a Real-time computer vision based ML model (Object detector) to detect status of traffic lights. Edge device will continuously stream the status of traffic lights to a in memory DB. Data model will also contain latitude & longitude of the traffic signal.

User Journey User will add source & destination location, we will pass it to Google Directions API & get the route. Based on the route we will fetch statuses of the traffic signals falling on the routes & through ML engine find out the speed that the user should maintain in order to pass the traffic lights when it is green.

Challenges we ran into

NA as we did not implement, ML model that would be required to solve such a problem will be similar to time window based Vehicle routing problem.

Accomplishments that we're proud of

The Idea itself as it create impact in following areas :

• Decreases Driver's Stress : Total on road vehicles are 295 Million , if we can help 50% drivers skip one traffic signal by showing him the optimal vehicle speed he/she should maintain and it impacts them by decreasing their stress by 1% ,Overall we would have decreased 1475000% stress from all our user's lives.
• Helps to decrease pollution
• Provides future scope for better first response team( Explained below)

What's next for No-Traff*cked

When all the city's traffic lights would have camera's installed we will be able to create a digital twin of the city's transport infrastructure & simulate city's traffic volume to predict which roads to upgrade, blind spots, accident prone areas & communicate with first res-ponder teams to call help if it happens within the camera frame !

Built With

• computer-vision
• fastapi
• googlecloudfunctions
• machine-learning
• pubsub
• python
• sql