CrowdCount

The number one passenger complaint is "I can't get a seat", but train operators don't know how many people actually board their trains

When planning, Train Operating Companies need to know: Can we adjust pricing to manage demand? If journeys are full for part of the route, should we run short-haul shuttles?

When booking, Passengers want to know:
Which trains are less crowded?

When at the station, Passengers want to know: Will I get a seat? Which coaches will be less crowded? Will the train after this one be less crowded? Should I get another coffee and wait?

Infrared (thermal) cameras have reduced dramatically in size and cost. This project trials a commercially available camera to determine its ability to identify and count boarding passengers. We also build an API to support passenger counting regardless of source to serve queries on historical data for planning and booking and to serve near real-time data for query by apps serving passengers in transit.

Built With

• android
• flir-thermal-camera
• ruby-on-rails