Académique Documents
Professionnel Documents
Culture Documents
Transport Benchmarking:
Learning from others
Ben Condry, Senior Research Associate
RTSC, Imperial College London
CoMET
Key elements of the presentation
2
Imperial College London and the RTSC
3
RTSC history and experience – 19 years of successful
worldwide benchmarking projects
Community of Metros
1994 Group of Five metros
CoMET
1996 Community of Metros (CoMET) for large metros
4
CoMET and Nova Metro benchmarking groups
Now 31 metros world-wide, including most of the largest
London DLR
Brussels Moscow
Newcastle
Berlin
Toronto London
Underground Paris Istanbul Beijing
Montreal
Naples Nanjing Shanghai
Lisbon Barcelona
New York Guangzhou
Madrid Taipei
Singapore
Rio de Janeiro
Sao Paulo
Sydney
Trains
Santiago
Buenos Aires
CoMET
Less than 500 million passengers p.a. More than 500 million passengers p.a
5
ISBeRG International Suburban Rail Benchmarking
Group 15 Railways, 14 Cities, 6 Continents
Oslo
(NSB)
Copenhagen
(S-Tog)
London
Munich
(Overground) Tokyo
(S-Bahn)
New York (JR East)
San Barcelona
Francisco (Long Island Rail Road, Hong Kong
(FGC)
(BART) Metro-North) (MTR)
Sydney Brisbane
(Sydney Trains) (Queensland Rail)
Sao Paulo
(CPTM)
Melbourne
Cape Town (PRASA)
(Metro Trains)
6
International Bus Benchmarking Group
12 Member Cities, 13 Operators
Dublin
Vancouver Brussels
Montreal Paris
Istanbul
Seattle London
Singapore
7
American Bus Benchmarking Group: 17 Members Across the
US in a Wide Range of Urban and Suburban Environments
Spokane Transit
(Spokane)
C-Tran
(Vancouver)
RGRTA
Lane Transit DART NFTA (Rochester)
(Eugene) (Des Moines) (Buffalo)
Pace GCRTA
UTA (Cleveland)
(Chicago) RIPTA
San Joaquin RTD (Salt Lake City)
(Rhode Island)
(Stockton) RTA
(Dayton)
Omnitrans
(San Bernardino)
The T
(Fort Worth)
Capital Metro
(Austin)
LYNX
(Orlando)
PSTA
(St. Petersburg)
8
Why do major transport operators find benchmarking so
valuable?
Public transport essential for cities to function effectively
Our operators spend over £40bn (70bn AUD) per
annum, using scarce public funds
Growing expectations demand modern, safe, reliable
and efficient networks.
Limited opportunities for operators to gauge
performance locally
Benchmarking is a key tool for operators to see if they
are operating optimally, and how to improve
9
Many common strategic and technical challenges -
Benchmarking has become essential to help address these
Top Strategic Challenges for Metros
10
For members, the benchmarking groups are central to
proactive, effective continuous improvement
Identify high
Why do others do priority issues
things differently? and areas for What has
Can we lean from improvement worked (or not)
them? elsewhere?
“The Search for Best Practices that
Lead to Superior Performance”
Best practice
Informed
transfer and
dialogue with
implementation
Set challenging but stakeholders
achievable
performance targets
11
Key principles of the benchmarking groups:
Confidentiality, Collaboration, Speed and Independence
12
The benchmarking process: a continuous annual cycle
80
60
40
20
10
13
KPI System – measures “whole system” performance
based on the Balanced Scorecard
Growth, Learning & Innovation Internal Processes
% change Network Size & Passenger Journeys Reliability & Availability
% change Operated Capacity km & Car km % of Cars Available & Used in Peak Hour
Number of Training Hours / 1000 Staff Hours Car km / hours between Incidents (by category)
Non-fare Commercial Revenue /Fare Revenue &
/Passenger Journey Efficiency
Passenger Journeys / Staff + Contractor hours
Financial Capacity & Car km / Staff + Contractor hours
Total Commercial Revenue / Operating Cost Train hours / Driver Hours
Operating Cost / Revenue Car km % Employee Absenteeism
Service Operations Cost / Car km Traction Energy Consumed / Car km
Maintenance Cost / Car km Total Energy Consumed / Passenger Journey & km
Administrative cost / Car km
Investment cost / Car km
Operating Cost / Passenger Journey & km Safety & Security
Fare Revenue / Passenger Journey & km Total Fatalities / Passenger Journeys
Deaths from Suicide / Passenger Journeys
Customer Deaths from Accidents / Passenger Journeys
Capacity Provision & Utilisation Deaths from Illegal Activity / Passenger Journeys
Capacity km / Route km Incidences of Crime / Passenger Journeys
Passenger km / Capacity km Staff Lost Time through Accidents / Staff Hours
Service Quality
Passenger Hours’ Delay / Passenger Journey Environment
Passenger Journeys On Time / Passenger Journey
CO2 per Passenger km
Trains On Time / Total Trains (scheduled + actual)
Train Hours Operated / Hours of Train Delay
14
Purpose and use of Key Performance Indicators
15
KPIs are just the starting point – the greatest value
comes from applying Best Practice leant from others
16
A KPI which describes a lot about benchmarking
Is such disparity in performance real?
Car km Between Incidents Causing a Delay > 5 Minutes to Service (2011)
Research Challenge: identify and quantify factors within and outside
management control that affect reliability
18
The numerical benchmarking has to be combined with
“Best Practice Benchmarking” – Drill-Down Studies
Almost 500 topics now covered
Drill-Down studies conducted annually in studies; examples include
by each group
Engineering
Members select topics of mutual •Rolling Stock Reliability
interest •Asset Renewal Decision Making
Reflect key topical issues, and questions
raised by KPIs Operations
•Investing in Punctuality
Members provide data and •Driver Productivity
Questionnaires and expert interviews
Customers
Imperial College analyses and reports •Customer Satisfaction
findings and recommendations •Passenger Information
19
Benchmarking Benefits:
Improved decision-making by learning from the experience of others
Asian Metro: driver productivity study: 10% saved through
shift reorganization
20
Example: escalator engineering benchmarking:
application with high impact
21
Some Benchmarking Examples -
Helping with Key Challenges
Community of Metros
CoMET
22
Railways/metros have returns to density and not scale
(so we can compare organisations of different sizes)
Network Size and Passenger Journeys p.a. (2011)
600 3 .0
Nova Metros CoMET Metros
500 2 .5
100 0 .5
0 0 .0
Network Length
700
Stations
600 Sydney Trains’
network still very Passenger Journeys
500 large in ISBeRG
400
300
200
100
0
Tk SP HK Sy Mb Mu CT LO SF Ch LI MN Bc Br Os
10
suburban railways,
2.5 Sydney is also relatively
low density
1.0
0.5
0.0
As Am As Eu Af Am Eu As Eu Eu Sy Am Am As Eu
27
Supply varies both spatially and temporally
(Here: Buses in major cities – over 24 hours)
% of total fleet
Use of Vehicles in Revenue Service (Average Weekday) - 2008
100%
90%
80%
70%
60%
50%
40%
30%
28
Key challenge that benchmarking addresses: off-setting wage
growth with revenue growth and increased productivity
Cost of Labour per Hour
(Nominal prices: Index 100 = earliest data year)
29
When economies were strong, a combination of poor fares policy and
rising costs set metros in the wrong financial direction
Real Labour
Fares Productivity
Falling Unit Falling Demand
60% Labour 75% up for
Metros & Energy Metros* 89% of
Costs Up metros
65%
Unit Energy
Real Fares
Costs
Falling
Labour Falling 60% Unit Labour
64% Metros
Productivity metros Costs
Improved Up
84% Metros 68% metros
31
On average, operating costs +40% has been spent on reinvestment
But fares revenue barely covers operating costs (except in Asia)
Commercial Revenue per Operating Cost, 2011
2.5
Other Commercial Revenue
Fare Revenue
Key factors: density, fares,
2.0 network, efficiency
Observed Metro
1.5 Reinvestment Rate
(~40%)
Revenue = operating cost
1.0
0.5
0.0
32
Case Studies drill down to understand ‘how’ and ‘why’
operators can become more productive (examples)
50/50 productivity gain shares
33
CoMET and Nova benchmarking shares experience in use of
technology to improve productivity / effectiveness
Unattended Train Operation (UTO) (Paris Line 1 & 14)
34
Factor of 9 variation in staff productivity: What is the right level? What is
achievable? Benchmarking helps transport providers make such choices
Car km per Total Staff and Contractor Hours
2 .0 More
Such high level KPIs can mask
1 .8
efficient?
efficiencies and inefficiencies in sub-
divisions: disaggregation,
1 .6 econometric modelling and drill-
down studies are performed
1 .4
Index (Mean = 1)
Less
1 .2
efficient?
1 .0
0 .8
Structural Factors: e.g. Density
0 .6 Quality (customer service etc)
0 .4
0 .2
0 .0
35
Continued world economic growth affects major capital cities such as
London, New York, driving continued rapid passenger growth
Annual Passenger Journeys % Change
3 .0 2006 - 2011
-4%
Growth continues in
2 .5 London, Paris, New Network expansion
York driving passenger growth
Billion Passenger Journeys
in Asian metros
2 .0
11% 10%
6%
70%
1 .5 123%
14%
119%
132%
1 .0 -2% 94% 44%
50%
48% 259%
0 .5
0%
0 .0
As As As Eu Am Eu Eu Am Eu Am As Am As As Eu As
2005 2006 2007 2008 2009 2010 2011
36
Demand for rail also growing across the world
550
1.5%
500 Major growth in London 3.2%
450 (network expansion) and Sao
400 Paulo (service improvements
350 and economic growth) 2.9%
300
2.4% 0.7%
250
200
0.4%
150 -0.8% -0.8% 0.8%
2.4% 0.6% 0.6%
100
0.6%
50
0
Eu As Eu As Am Eu As Am Eu Eu Am Am As As
2007 2008 2009 2010 2011 2012
Average Annual GDP Growth Over Equivalent Time Period 37
Conclusions: Benchmarking can be highly effective
38
Any questions?
b.condry@imperial.ac.uk
www.imperial.ac.uk/rtsc
39
Metro Codes – used in graphs to represent metros
Community of Metros
CoMET
BA – Buenos Aires Metrovías Bg – BMTROC, Beijing
Bc – Barcelona TMB Bn – BVG, Berlin
Bs – Brussels STIB Gz – Guangzhou Metro Corporation
Bk – Bangkok BMCL HK – MTRC, Hong Kong
Dh – Delhi Metro Rail Corporation Ln – LUL, London
Is – Istanbul Ulasim MC – STC, Mexico City
Lb – Lisbon Metropolitano de Lisboa Md – Metro de Madrid, Madrid
Mt – Montréal STM Mw – MoM, Moscow
Nc – Newcastle Nexus NY – NYCT, New York
Nj – Nanjing Metro Pm – RATP Metro, Paris
Np – Naples Metronapoli Pr – RATP RER, Paris
RJ – Rio Metro SC – Metro de Santiago
Sg – Singapore SMRT Sh – SSMG, Shanghai
Sy – Sydney Trains SP – MSP, São Paulo
To – Toronto TTC Tp – Taipei TRTC