TfL is conducting a four week trial, collecting depersonalised (pseudonimised) WiFi connection data in order to better understand how London Underground passengers move through stations and interchange between lines. The data will be collected at 54 Tube stations within Zones 1-4, and if the trial is successful the intention is that the data could be used to improve services, provide better travel information and help prioritise investment across the Tube network. 

TfL is conducting a short trial, collecting WiFi data from 54 London Underground stations for four weeks

TfL is conducting a short trial, collecting WiFi data from 54 London Underground stations for four weeks. Details are on posters in stations.

WiFi data trial is underway

The WiFi data trial will last for four weeks from 21 November and will help give TfL a more accurate understanding of how people move through stations, interchange between services and how crowding develops. By analysing the in-station WiFi connection data, a number of potential benefits have been identified:

  • Providing better customer information for journey planning and avoiding congestion
  • Helping TfL better manage disruptions and events and ensure a safe environment for all
  • Better planning of timetables, station designs and major station upgrades

By understanding how customers move through and around stations, TfL also believes it may be able increase revenue from companies who advertise on poster sites or rent retail units, and this revenue would be used to reinvest in improving services across London.

“This short trial will help us understand whether WiFi connection data could help us plan and operate our transport network more effectively for customers. Historically, if we wanted to know how people travelled we would have to rely on paper surveys and manual counting, which is expensive, time consuming and limited in detail and reliability. We hope the results of this trial will enable us to provide customers with even better information for journey planning and avoiding congestion.”
Shashi Verma, Chief Technology Officer at Transport for London

How it works

The trial will work by collecting WiFi connection requests from mobile devices as customers pass through stations. When a device has WiFi enabled, it will continually search for a WiFi network by sending out a unique identifier – known as a Media Access Control address – to nearby routers. The data collected is automatically de-personalised (pseudonimised), which means that no browsing data will be collected and TfL will not be able to identify any individuals.

sdvsdvsdfv

The poster in Tube stations lets passengers know about the WiFi trial

No data collected through the trial will be made available to any third-parties, and posters will be on display within stations to let customers know that the trial is taking place. Should any customers wish to opt-out of the trial, they simply need to turn off their WiFi while passing through the station.

TfL already uses a range of data, such as aggregated and de-personalised Oyster and Contactless payment data and manual paper surveys, in order to understand how customers travel across London. While these data sources provide detail on the origin and destination of customer journeys, there are many options that customers could take across the network to complete their journey.

Traditional paper surveys are also expensive, take time to process and can only provide a snapshot of travel patterns on the day of survey. They are also unable to provide the continuous information detailing the varied travel patterns on the network.

“TfL is unlocking the power of data to gain insights into how passengers are using the network and drive its transformation into a smart transport system. The availability of big data analytics tools and technologies means that organisations, of all sizes and sectors, are increasingly able to make data driven decisions that can make a real difference to customers’ lives. In this case, it will mean more accurate passenger insights and easier journeys for customers.”
Sue Daley, Head of Big Data, Cloud & Mobile at techUK

For more information about the trial, please visit www.tfl.gov.uk/privacy

Posted by Stephen Irvine

Stephen is the Community Manager and Digital Blog editor for Transport for London

13 Comments

  1. […] Source: Wifi Data Trial – Understanding London Underground Customer Journeys […]

    Like

    Reply

  2. This is a great idea – shame about not letting us have the data, of course!

    Having a chat with the Hackathon team who were trying out Wifi for ticketing, I realised that the above mentioned collection of Wifi data would be very, very, very useful for prioritising the revenue protection people: you could tell in real time the differences between Oyster/Contactless charges and the Wifi data.

    No mention of that on the posters. Unless that’s “help prioritise investment”?

    Like

    Reply

    1. Hi Brian. Using Wifi for revenue protection purposes is not something that is in scope of the pilot. As this is sensitive customer data, TfL would have to take in to careful consideration how they use this data in the future and we would need to work in close proximity with the Information Commissioners Office (ICO). We do not currently have the capacity or capability to process this data in ‘real time’ so any processing of pseudonymised data is done so historically.

      Like

      Reply

  3. Stephen,

    Thanks for the extra details. I think this project is a great idea – it’s worked well for planning crowds at football stadiums I understand.

    How granular will the Wifi details be?

    Presuming you are using the existing infrastructure, will it only be station-level or will you know more accurate locations based on the router, power level and aerial direction (where relevant)?

    Is this just the “Virgin Media Wifi” or will it also use other TfL wifi routers such as those for London Overground and Crossrail?

    When testing the Crossrail Staff App, it was possible to use the combination of Apple iOS Wifi services and the available Wifi networks to detect the correct TfL Rail station – the iPad minis they use don’t have GPS.

    Like

    Reply

    1. Yes, we’re using our existing infrastructure so we are only able to identify the router location. The information that we are collecting from this pilot at 54 distinct underground stations across the network, and the routers within these stations have unique identifiers which is necessary to answer questions that we have set for the pilot to see if WiFi is a useful data source. These questions are,

      · Operations and safety information

      Understanding how customers move around stations could help us to manage disruptions and events more effectively, deploy staff to best meet customer needs and ensure a safe environment for all.

      · Transport planning

      By better understanding how our customers use the Underground network, we can better plan timetables, station designs and major station upgrades.

      Liked by 1 person

      Reply

  4. Will this trial be able to identify *where* in the station the user is?

    I’m thinking for the future that an app might want to know that a user is in carriage 2, the train is coming in on platform 3 and their next train leaves from platform 7. Then it can tell the user whether to turn left or right to reach the best stairs or escalators and then keep them updated as they pass through different corridors. All could be more efficient than looking for signs and having to read destination boards.

    Liked by 1 person

    Reply

    1. Hi Harry. The data collected is able to see which router a device is connected to, the accuracy of this will change depending on the layout of the station and the number of routers within a station. We do not currently have the capacity or capability to process this data in ‘real time’ so any processing of pseudonymised data is done so historically.

      Like

      Reply

  5. Very interesting project!
    I wrote my PhD thesis about collecting and modelling WiFi data. It was quite similar to this context. I just share here few pointers, in case it might be useful:
    – a 2-minute video about the full project: https://www.youtube.com/watch?v=IckZq2wGzwQ
    – my PhD thesis (full text): http://dx.doi.org/10.5075/epfl-thesis-6806
    – the WiFi data I used on EPFL Campus: http://doi.org/10.5281/zenodo.15798 for reuse
    – a paper about how to merge WiFi data with occupancy data and map info/land use: https://infoscience.epfl.ch/record/199471 (note that there is a free postprint, if you don’t have access to the journal)
    – a second paper about how to use WiFi data in destination choice models (these are panel data, and it creates small challenges for data analysis): http://dx.doi.org/10.1016/j.jocm.2016.04.003 (published with data and models. No free version here, but audioslides are here: https://www.youtube.com/watch?v=CZHV_klsCK0 and an old version of this paper is in my thesis, as well as (old) data and models here: http://doi.org/10.5281/zenodo.33973)
    Enjoy watching/reading 🙂

    Liked by 2 people

    Reply

    1. Hi Antonin. Thank you for this information, I will pass this on to our operational research team, and I’m sure this is something that will be of interest to them.

      Liked by 1 person

      Reply

  6. An idea: trial releasing real-time data on number of WiFi connections within stations as an indicator of how busy they are. Even better, since you know which areas of the station people are connected in, give an estimate of how busy platforms are.

    It’s a missing piece in planning an efficient, bearable journey. And it doesn’t require tracking individual people. Or is some equivalent data already public?

    Like

    Reply

    1. We do not currently have the capacity or capability to process this data in ‘real time’ so any processing of pseudonymised data is done so historically and this pilot is looking to see if the wifi data can help wiht the following, which is similar to your question:

      “Operations and safety information – Understanding how customers move around stations could help us to manage disruptions and events more effectively, deploy staff to best meet customer needs and ensure a safe environment for all.”

      Like

      Reply

  7. Here’s What TfL Learned From Tracking Your Phone On the Tube – By James O Malley on 13 Feb 2017 at 12:24PM

    http://www.gizmodo.co.uk/2017/02/heres-what-tfl-learned-from-tracking-your-phone-on-the-tube/

    “Today, thanks to the Freedom of Information Act, Gizmodo UK can exclusively reveal some of the utterly fascinating findings that the agency has been able to make from all of our data – and how the plan, if the trial is deemed a success and tracking is implemented full time, is also to use the data to inform advertising decisions on the Tube network.”

    Like

    Reply

  8. […] confined to heritage sites. It has been implemented in airports, shopping centres, and even on the London Underground. Transport for London use the software to look into how they can improve their services. For […]

    Like

    Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s