1

Understanding people on the move in London

Charles BuckinghamImpacts Monitoring Manager

(with thanks to….,.)

Lauren Sager WeinsteinHead of Oyster Development

Contents

•

TfL’s responsibilities•

About Oyster

•

Using Oyster to understand people on the move

•

Congestion Charging –

some examples

2

About TfL

•

Mayor of London’s transport authority

•

Finances/procures/operates/maintains public transportation

–

London Underground

–

Buses

–

Docklands Light Railway

–

Croydon Tram

–

TfL Road Network – 580km of arterial roads

–

Congestion Charging

3

TfL’s responsibilities

4

TfL travel facts

Every weekday in Greater London:

6 million journeys are made on London’s buses

3.4 million on the Tube

11 million car / motorcycle trips

155k + passengers on DLR

9.5 million walking or cycling trips

70% of National Rail journeys begin or end in London

5

About Oyster

Source: TfL Fares & Ticketing Directorate

•

TfL’s multi-modal smartcard–

National rail–

London Underground–

Buses–

Tram–

DLR•

Contactless: 0.2s read/write at the reader•

3 x tickets + £90 PAYG with daily capping•

Distributed to customers free with a period travel product or a £3 returnable deposit: >16m issued to date

•

Concession & discount variants–

Freedom Pass for over-60s–

Various child & student schemes–

Bus & tram adult discount card

6

Penetration of Oyster

2004 2005 2006 2007

Source: TfL Fares & Ticketing Directorate

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2004 2005 2006 2007

Oyster share of all TfL trips

Eliminated magnetic 7-

day bus passesEliminated TfL

magnetic weekly

TravelCards

(TCs)

Reduced Oyster

PAYG fares relative to

cash

Oyster PAYG

introduced

7

Key Oyster Benefits

Change in customer behaviour

Old: purchase a ticket and then travel

New: streamlined travel for customer

Reduces queues

Minimises cash handling

Tackles fraud

Speeds customers through gate

Source: TfL Fares & Ticketing Directorate

8

Using Oyster to Understand travelling behaviour

Key Oyster statistics

• As of January 2008, 17m + Oyster cards issued• 5.6m cards were in use during the previous 4 week period

•

During the week of 25 November -1 December 2007, on an average weekday there were:

3.1 million Oyster journeys a day on the Tube and DLR5.4 million Oyster journeys a day on buses and trams

•

In November 2007, Oyster card journeys represented around 74% of bus and Tube journeys.

9

Oyster Card Personal Data

• Oyster cards can be registered or unregistered

• Registered cards can be protected if lost or stolen

• Mandatory registration on monthly and annual tickets

•

Detailed journey history kept for 8 weeks for customer service purposes (eg

refunds)

• After 8 weeks, personal data is anonymised

10

Understanding travel patterns using anonymised

Oyster data

Analysis work supported through TfL partnership with MIT, with TfL guidance on crafting research questions

Sample research

11

Using Oyster to measure Variation of OD Journey Time

Oyster Journey 

Time Percentile

10th 25th 50th 75th 90th

Minutes

Range of 

journey time 

experienced by 

the middle 

80% of 

passengers

Range of 

journey time 

experienced 

by the middle 

50% of 

passengers

•

Example ranges only

12

Can we use Oyster data to measure variability of journeys between stations?Research by Joanne Chan, MIT MST 2007

Results – Victoria Line (AM Peak Northbound)

Origin Stations to All Northbound Destinations

•

Skewed distribution•

Victoria, the only Zone 1 station in the graph–

Largest average excess minutes–

Largest variation in excess minutes

-1

0

1

2

3

4

5

Exce

ss M

inut

es

PC90

PC75

PC50

PC25

PC10

AVG

Brixton

(4,277)

Vauxhall

(3,365)

Victoria

(2,780)

Finsbury

Park

(287)# of Oyster Observations

Research by Joanne Chan, MIT MST 2007

Using Oyster to Measure Crowding

Can we use Oyster data to capture effects of crowding?Research by David Uniman, MST candidate 2008 The theoretical model:•

Imbalance b/w Travel Demand ↔ Transport Supply–

Platform Crowding

–

On-Train Crowding–

In-Station Crowding

•

Leads to Increased User Travel Times–

↑

Wait Times

–

↑

On-Train Times•

Through…–

Full Trains

–

Headway Variations

Dwell Times

Schedule Adherence

Platform Crowding

Full Trains14

Analysis: 30-min AM Peak Oyster

Victoria --> Oxford Circus

4

5

6

7

8

9

10

11

12

13

14

7:00-7:30am(54)

7:30-8am(134)

8-8:30am(460)

8:30-9am(315)

9-9:30am(296)

9:30-10am(180)

AM Peak

TT [m

in]

10th Percentile25th Percentile50th Percentile75th Percentile90th PercentileRP Corr2

RP Corr2 = TfL rail plan modelling tool, corrected to take into account Oyster journeys are gateline

to gateline

David Uniman, MST Candidate MIT 2008

Travel time increases at the peak; after the peak many journeys are still longer than early morning

15

16

Research Question

•

By focusing on bus passenger interchange behaviour, can Oyster data be used to help improve the public transport network in London?

•

Key contribution:–

Methodology for describing passenger interchange behaviour in London using Oyster card data

Catherine Seaborn, MIT MST candidate 2008

17

Journey Segments Per Passenger

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

40.0%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Perc

enta

ge o

f Tot

al P

asse

nger

s

Number of Daily Journey Segments

Daily Journey Segments Per PassengerAll Oyster Card Modes

Source: 5% Oyster data for 2007 Period 2 (April 29 – May 26)

17.1%

34.9%

14.7% 14.1%

7.0%

4.8%

2.7%1.8%

1.1% 0.7% 0.4% 0.3% 0.1%0.2% 0.1%

•

Question: how many journey segments do Oyster customers take on

a given day?

Catherine Seaborn, MIT MST Candidate 2008

18

Weekday Journey Segment Patterns

Mode 1 Mode 2 Mode 3 Mode 4 Mode 5 Mode 6 Passengers ShareCumulative Share

U U 416,082 16.3% 16.3%

B B 401,356 15.7% 32.0%

B 266,561 10.4% 42.4%

B B B 150,781 5.9% 48.3%

B B B B 144,275 5.6% 54.0%

U 125,528 4.9% 58.9%

B U U B 77,353 3.0% 61.9%

B B B B B 72,943 2.9% 64.8%

U U U 65,190 2.6% 67.3%

B B B B B B 50,485 2.0% 69.3%

Source: 100% Oyster data for Wednesday, November 14, 2007

•

What are the modes for these journey segments?

•

Top 10 shown

•

Total patterns: 15,802

Catherine Seaborn, MIT MST Candidate 2008

19

Research Question

•

What are the characteristics of interchanges to bus at London Underground/bus interchange locations?–

How long does it take for passengers to transfer between modes?

–

Function of walk time, frequency of service, reliability

Catherine Seaborn, MIT MST Candidate 2008

Catherine Seaborn10 January 2008

20

Example Interchange Stations

21

Potential Interchange Time: Underground-Bus

Time Difference Between Underground Station Exit and Bus BoardingAll Oyster Card Passengers, Single Weekday

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 5 10 15 20 25 30 35 40 45 50 55

Time from Station Exit to Bus Boarding (minutes)

Perc

enta

ge o

f Pas

seng

ers

Boa

rdin

g B

uses

With

in 6

0 M

inut

es

.

Oxford Circus (38)Angel (19)Holloway Road (16)Burnt Oak (11)Northwood Hills (3)

Source: 100% Oyster data for Wednesday, November 14, 2007

Note: Number denotes total bus routes serving station, including night buses.

Catherine Seaborn, MIT MST Candidate 2008

Congestion charging in central London

22

Extended Central London charging zone

23

A transport success

24

•

Traffic entering charging zone: (4+ wheels)

Down 21%

•

Chargeable vehicles: Down 31%

•

Initial impact on congestion high: 30% decline (first yr) Averaging 21% over scheme lifetime

•

Nitrogen oxides (NOx

) emissions: Down 13% 8% due to Congestion Charging

•

Particulate matter (PM10

) emissions: Down 16%6% due to Congestion Charging

•

Carbon Dioxide (CO2

) emissions: Down 16%

Substantial traffic change

25

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

180,000

200,000

Cars and minicabs

Vans Lorries and others

Taxis Buses and coaches

Powered two-

wheelers

Pedal cycles

Total flow

Spring 2002Autumn 2002January 2003Feb/Mar 2003Spring 2003Autumn 2003Spring 2004Autumn 2004March 2005Spring 2005Autumn 2005November 2005Spring 2006Autumn 2006Spring 2007

Camera-based enforcement (1)

26

Camera-based enforcement (2)

27

Colour Contextual ImageColour Contextual Image

Monochrome Image from ANPR cameraMonochrome Image from ANPR camera

ANPR system outputANPR system output

Evidential Record SummaryEvidential Record Summary

Number Plate image from ANPRcamera, Lane 1

Number Plate image from ANPRcamera, Lane 1

Unique opportunity to study traffic characteristics and behavioural change

28

Potent data source:

Vehicle population profilesFrequency of travel etc.Some routeing/journey time information (congestion)Match with licensing data –

vehicles registered not same as vehicles

‘in the zone’

BUT:

Cameras capture vehicles NOT peopleOnly captures vehicles ‘there’

–

not those who have gone away

Do not capture whole tripData Protection imposes some (necessary) limitationsCan’t really use as sample frame for follow-on surveys Cameras optimised for enforcement NOT researchTend to be defeated by ‘easier’

things like data processing

So:Potential only partially fulfilled

Understanding our chargepayers

29

Measuring congestion

30

0.0

0.5

1.0

1.5

2.0

2.5

3.0

MarApr

MayJun

JulAug

SepOct

NovDec

JanFeb

MarApr

MayJun

MarApr

MayJun

SepOct

MayJun

NovDec

Exce

ss d

elay

(min

/km

)

Automatic number plate reading camera dataMoving car survey data

2003 2004 2005 2006

Understanding the effects of charging

31

Potential future developments - Tag & Beacon

Tag and beacon technology is already providing high capture rates for schemes where charges vary across the day, for example cordon charging varying by time of day

Stockholm 2006

The future? Satellite/mobile positioning systems

Satellite and mobile phone location systems for distance-based charging need further development for affordable use in urban areas

Example position reports from multiple different GPS and GSM mobile devices

Example from GPS trials - ‘use’ of zone varies by time of first entry

34

Thank You !

35

www.tfl.gov.uk