UrbanVind

Data collection in public transportation is a challenge in Europe. Cities struggle to gather crucial data for operations and planning. Information such as where people board and exit and how many passengers are sitting on a bus is not readily or easily available, as was found after interviewing 40+ cities and operators. Traditional solutions to count passengers, like Automatic Passenger Counters, are expensive and maintenance-heavy and do not live up to their claimed accuracy. Techniques for determining travel patterns are often inaccurate because cities only know where a passenger boards.

UrbanVind’s CrowdFlow offers a solution: accurate, easy-to-deploy, cost-effective software providing real-time vehicle occupancy and passenger flow data. This data is not only useful for a bus operators and planners, but can be used to inform passengers of crowding in public transport and is an enabler for predictive AI tools that optimise public transport.

CrowdFlow uses Bluetooth to detect nearby devices and estimate passenger numbers. This estimation model, tailored to each city, delivers real-time data to public transport authorities. By tracking Bluetooth devices, accurate OD Matrices are generated. Currently in beta, CrowdFlow shows around 80% accuracy. Antalya, Türkiye and Jurmala, Latvia need better vehicle occupancy and passenger flow data and real-time crowding insights to improve their day-to-day operations and to build a better and more accessible network. These two cities will implement CrowdFlow across their bus networks with the goal of bringing Crowdflow to commercial readiness so that it can support more cities across Europe to improve their public transport networks.