Joined the team
Real-time occupancy rate measurement in public transportation
Real time occupancy measurment in the public transport can be done through various ways. We chose to go with an AI based approach, with person recognition, running on a low cost hardware.
Network and power are not major concerns as the bus has wired power and 3G connection onboard.
Our setup based on YoloV3/ImageNet running on a raspberry Pi, with a GPS shield can analyze realtime and send the numbers to our backend, and show the infos through an Angular dashboard, both running in Heroku.