/ CALL US + 44 (0) 20 3873 0260
  • SmartTransit Congres, Boston - 2020

How TfL / London Underground is Wi-Fi tracking its users (Part 1)

Posted by Marcello Perricone on Sep 2, 2019

tfl tube train pn056 - S Stock at Hammersmith

Information is power. From warfare to business, the ability to efficiently gather knowledge and understand it broadens horizons, expands options, and makes possible what seemed impossible. As technology inexorably advances day by day, the capabilities to acquire and process data invariably follow, and its innovative uses can change the way people move about in the world.

In May 2019. Transport for London (TfL) announced a project straight out of a science fiction novel -- London Underground's free Wi-Fi systems would start tracking users throughout stations, analysing their data anonymously to better understand how passengers used the transit system and upgrade the network accordingly.

The project -- built upon the Tube's existing Wi-Fi network -- started with a pilot program back in 2016 to test the viability of the concept.

"Our Wi-Fi data collection project has been something that we have been looking into for several years," says Lauren Sager Weinstein, Chief Data Officer at TfL. "Currently, TfL uses data from its ticketing system to understand how journeys are made across the network. While this is accurate for people entering and exiting the stations, this data cannot show the flow of movement through a station. So in 2016, we held a four-week long pilot to test whether we could use depersonalised data from our in-station Wi-Fi network collection to better understand patterns of how groups of people move across our network."

Lauren Sager Weinstein"When a device has Wi-Fi enabled, it will continually search for a Wi-Fi network by sending out a unique identifier - known as a Media Access Control address – to nearby routers as customers pass through stations. The pilot ran across 54 stations within Zones 1-4 and collected these Wi-Fi connection requests, which were automatically depersonalised," says Lauren. "Our in-house analytics team then carried out analysis to help understand where customers were at particular points of their journeys. During the four week pilot, more than 509 million depersonalised pieces of data were collected from 5.6 million mobile devices making around 42 million journeys, which revealed a number of results to TfL that could not have been detected from ticketing data or paper-based surveys."

The success of the pilot program back in 2016 proved that the data could not only be gathered and analysed efficiently, but also avoid the moral quandaries of tracking users by depersonalising information to avoid identifying individuals. This last aspect was crucial to the full implementation of project, as TfL was well aware of the fallout should the tracking initiative breached customer's privacy and backfired.

"Since the pilot, we used the ideas and lessons gathered to work with our colleagues across TfL so that once we started collecting, we could process and present the information in ways that our customers and operational teams would find helpful and clear to understand," explains Lauren. "To ensure that customer privacy was at the heart of the project, we also worked closely with key stakeholders and the Information Commissioner's Office to follow ‘Privacy by Design’ principles, so that privacy concerns and transparency were actively considered and addressed."

tfl infographic PUB17_031 WiFi_pg21fig

In order to fully roll out the project in 2019, Transport for London undertook a massive modernisation effort in the intervening 3 years, going as far as creating 3D models of every single Tube station in the network.

"While the technology is built on the back of our existing Wi-Fi network, a huge amount of work was required to get to the system we have now introduced. Detailed digital mapping of all London Underground stations was also undertaken to allow TfL to identify where Wi-Fi routers are located and allow TfL to understand in detail how people move across the network and through stations," says Lauren. "When we ran the pilot, we turned the system on and collected the data during the whole course of the pilot. However, we only started analysing the data once we turned off data collection and started to analyse the pseudonimysed ‎data. A full scale introduction across the network however would have to work differently."

"In order to understand the complex movement patterns, we needed to have detailed station mapping of where all of the Wi-Fi access points are so that we could assign the connection to a specific location. So we first of all undertook a very extensive mapping exercise to make sure that all our access points were reflected in our station models and, when there are updates, we have a process to ensure that these models are kept up to date. This includes taking into account where the Wi-Fi access points are in relation to corridors, lifts, escalators, etc."

"We also had to think about how this depersonalised and aggregated data could provide value to our operational teams," continues Lauren. "We therefore worked with them to identify baseline profiles for stations, and then designed algorithms to work in real time. These can then highlight particularly busy times and places, which will benefit both for our operational controls and allow us to directly inform our customers to consider alternative routes or retime their journey."

This is part 1 of a two-part feature. To find out how TfL went from a small pilot program to a fully implemented, full scale Wi-Fi tracking of its passengers throughout its whole network of 270 stations, read part 2 here.

New call-to-action

To meet some of TfL's top executives, experts, and decision makers, book your ticket now and join us at SmartMetro Madrid, on November 25-27th!

Topics: Transport For London (TfL), Wi-Fi, SRW Featured

Marcello Perricone

Written by Marcello Perricone

The Editor of SmartRail World and Transport Security World.

Get The Latest Updates From SmartRail World

Please use the form below to leave a comment about this story.