How TfL uses ‘big data’ to plan transport services
20 June 2016 • Author(s): Lauren Sager Weinstein, Head of Analytics at Transport for London (TfL)
Lauren Sager Weinstein, Head of Analytics at Transport for London (TfL), has responsibility for the analysis of customer data, supporting operational and planning areas in delivery of services to TfL’s customers. London is a big growing city; more than 31 million journeys are made in the capital each day, 23% more than 15 years ago. Trains on one of the busiest underground lines, the Victoria line, carries thousands of commuters at a rate of one every 100 seconds during the morning peak. As Lauren explains for Eurotransport, it’s therefore natural that the data that TfL has access to is big as well.
Every day there are 20 million ‘taps’ captured through our ticketing system; our iBus location system provides accurate location and prediction information for all 9,200 vehicles in the fleet and we help keep London moving by managing traffic flow with our 6,000 traffic signals and 1,400 cameras. Our systems therefore record a vast amount of operational data.
However, just holding lots of data isn’t enough. To get value from it we have to turn it into useful information for our customers and into tools to plan and run our services. We actively experiment to see what we can learn from all data. The results of these trials allow us to improve the products and services that we deliver to our customers.
Using our ticketing data we have been able to build a comprehensive picture of travel patterns across our rail and bus networks. The use of Oyster and contactless payments through bank cards, Apple Pay and now Android Pay, has given us tube and rail station entry and exit data as customers have to touch in and out for their journeys. Bus journeys, on the other hand, may seem more problematic to monitor, as our customers are only required to tap in when they get on, but not when they exit. However, we can now tell when our customers are leaving a bus using a Big Data tool which looks at origin, destination and bus interchange information – which we call ODX. It combines bus location and ticketing data to try and match up origin and destination pairs to create a multi-modal travel dataset.
All of this information means that we can improve network and interchange planning and review the impacts of closures and diversions. For example, we used ODX to restructure the bus network in the New Addington area of London to help provide better services for local residents. In October 2015 we launched a new service pattern for the neighbourhood that better meets our customers’ needs.