This talk by Professor Felix Schmid discussed a systematic approach to the analysis of railway system complexity, with a particular focus on interfaces and interactions. Case studies related to conventional railways, high-speed railways, light rail transit and novel technologies. For more information: http://goo.gl/qLtzY

1. UNIVERSITYOF
BIRMINGHAM
Determinants of Railway
System Complexity and
their Management
Dr. Felix Schmid
The University of Birmingham, in Association with
Armitage, Harris, McKechnie, Perrow, Qurashi, Reason

2. Determinants of Railway System Complexity and their Management
Typical Railways?
UNIVERSITYOF
Slide No: 2 University of Wollongong, 2012-22-23 BIRMINGHAM

3. Determinants of Railway System Complexity and their Management
Overview of Presentation
• What are determinants? A definition:
– Determinants are factors or issues that have the potential to
influence a decision, situation or context:
– Determinants are normally co-acting, i.e., there is rarely ever
a single determinant.
• A railway complexity case study;
• Review of railways’ natural characteristics;
• Managing the complexity of the railway system;
• A high-speed rail example of complexity management;
• Conclusion and discussion.
UNIVERSITYOF
Slide No: 3 University of Wollongong, 2012-22-23 BIRMINGHAM

4. Determinants of Railway System Complexity and their Management
Background to Complexity Case Study
• Birmingham, Manchester, Sheffield and Felix Schmid;
• A rather basic case study:
– The night of Friday, 7 January 2005 is dark and stormy;
– In the morning of Saturday, 8 January 2005, the storm is still
raging;
– John Hebblethwaite (not his real name) is rostered to take
the 05:11 from Sheffield to Manchester Airport and back…
– It is his first day at work after 6 months of sick leave;
– Events take a rather unpleasant turn for many;
– Case study lessons for the parties involved.
• The railway’s relationship with its competitors.
UNIVERSITYOF
Slide No: 4 University of Wollongong, 2012-22-23 BIRMINGHAM

5. Determinants of Railway System Complexity and their Management
Geography of Britain
Scotland
West Coast Main Line (WCML)
Glasgow Edinburgh
Carlisle
Blackpool Preston
Liverpool
Holyhead Sheffield
Crewe Manchester
WalesBirmingham
Fishguard
England
Rugby
Cardiff Bristol
LONDON
Penzance
UNIVERSITYOF
Slide No: 5 University of Wollongong, 2012-22-23 BIRMINGHAM

6. Determinants of Railway System Complexity and their Management
Current Roles of Felix Schmid
• Designer and director of MSc in Railway Systems
Engineering at University of Sheffield (1994-2008);
• Professor of Railway Systems Integration, University
of Birmingham (since 2005);
– Director of Education, Birmingham Centre for Railway
Research and Education;
– Leader of railway systems engineering courses for industry;
– Visiting lecturer at École Nationale des Ponts.
• Research in railway control systems & human factors;
• Expert on CrossRail’s systems engineering panel;
• Consultant to MWH and Crossrail for track systems.
UNIVERSITYOF
Slide No: 6 University of Wollongong, 2012-22-23 BIRMINGHAM

7. Determinants of Railway System Complexity and their Management
The World’s Most Wonderful Commute
Stockport
Sheffield M.
Chinley R2
R1 Chesterfield
UNIVERSITYOF
Slide No: 7 University of Wollongong, 2012-22-23 BIRMINGHAM

8. Determinants of Railway System Complexity and their Management
A Functional View of Railways
A fun
Standards more ctional re
usefu prese
Rules and
l than ntatio
Regulations
based a com n is
Laws analy pone
sis nt
Transport People
or Goods by Rail Transport Units Achieved (TUA)
Demand for
Transport by Transport Quality Achieved (TQA)
Railway
Transport Value Achieved (TVA)
/
rney ry
e jou cessa
es, th a ne
Funds
s
Resources &
o st ca ity is t valu
e
Equipment
In m rt activ eren
People
po o inh
trans and of n
evil UNIVERSITYOF
Slide No: 8 University of Wollongong, 2012-22-23 BIRMINGHAM

9. Determinants of Railway System Complexity and their Management
A ‘normal’ Train Journey to Work?
• Friday, January, 7 January
2005, 22.10: Forecast of
severe gales for Saturday;
• 8 January 2005: Leave home
at 06.35 to cycle to station –
in record time thanks to
strong following wind!
• Board 07:18 train to travel to
work in Sheffield, the former
home of the MSc in Railway
Systems Engineering;
• No problems apparent – train A Class 158 2-car Diesel Multiple Unit
leaves on time. waits to leave Piccadilly
UNIVERSITYOF
Slide No: 9 University of Wollongong, 2012-22-23 BIRMINGHAM

10. Determinants of Railway System Complexity and their Management
A Class 323 Train of Northern Rail
• ‘Northern Rail’ is franchise;
• Modern type of suburban
train found in Manchester
and West Midlands area;
• Largely used on regional APEX
commuter routes; FRAME
• 3-car electric multiple unit UPPER ARM PANTOGRAPH
HEAD
(25kV) powered by 3-phase CONTROL
ROD
induction motors;
KNUCKLE
LOWER
• Just one single arm high-
ARM
RAISING
CYLINDER
speed pantograph with low
AIR FEED INSULATOR
4th BAR
contact force;
• ‘My’ train is diesel multiple
unit of Class 158, built
around 1990, still quite new.
AIR
UNIVERSITYOF
EQUIPMENT
Slide No: 10 University of Wollongong, 2012-22-23INSULATORS
BASE & BIRMINGHAM

11. Determinants of Railway System Complexity and their Management
Determinant: Railway Diversity
The Number of distinct and
different sub-activities that are
performed within an integrated UNIVERSITYOF
Slide No: 11 system of tasks Wollongong, 2012-22-23
University of (McKechnie). BIRMINGHAM

12. Determinants of Railway System Complexity and their Management
Range of Railway Subsystems
Electrification & Traction &
Vehicle
Power Supplies Braking
Operations Management Structures
Systems
Maintenance System
Bogie
Communi- VCS
Systems
Control
cations &
Axles & Wheels
Signalling
Systems
CIS Rail ATP Rail
Sleepers & Ballast
Station Systems
Substructure System
CIS: Customer Information Systems / VCS: Vehicle Control Systems
UNIVERSITYOF
Slide No: 12 University of Wollongong, 2012-22-23 BIRMINGHAM

13. Determinants of Railway System Complexity and their Management
Customer and Passenger Needs Diversity
UNIVERSITYOF
Slide No: 13 University of Wollongong, 2012-22-23 BIRMINGHAM

14. Determinants of Railway System Complexity and their Management
Subsystem and Component Diversity
• Types of Subsystems:
– Switches and crossings;
– Electrification equipment;
– Power supplies & substations;
VH – Train control and signalling;
– Rolling stock and traction.
H • Component variety:
– Steel and concrete structures;
M – Microprocessors;
– Sensors and effectors;
L – Thyristors, GTOs, IGBTs;
–
Williams, ORR, 2006
Precision mechanical systems;
– Electrical machines.
• Issue raised by McKechnie
as ‘heterogeneïty’.
UNIVERSITYOF
Slide No: 14 University of Wollongong, 2012-22-23 BIRMINGHAM

15. UNIVERSITYOF
BIRMINGHAM
Railway are Diverse or
Heterogenous Systems

16. Determinants of Railway System Complexity and their Management
158 motors to Stockport and all is well
& some nice OHLE too!
A useful facing crossover!
UNIVERSITYOF
Slide No: 16 University of Wollongong, 2012-22-23 BIRMINGHAM

17. Determinants of Railway System Complexity and their Management
On Time in Stockport, depart after 60 s
Stockport Viaduct
on Normal Route
Edgeley Side of
UNIVERSITYOF
Slide No: 17 University of Wollongong, 2012-22-23
Stockport
BIRMINGHAM

18. Determinants of Railway System Complexity and their Management
Determinant: Railway Dispersion
Extent to which assets, resources and
people required for correct operation
of system are distributed over a large UNIVERSITYOF
Slide No: 18 area / along corridors (Schmid)
University of Wollongong, 2012-22-23 BIRMINGHAM

19. Determinants of Railway System Complexity and their Management
Dispersion and Linearity
• Linear infrastructures:
– 10 m wide and 1000s km long;
– Great impacts on environment;
– Environmental impact varies.
• Distributed assets:
– Assets difficult to reach;
– Assets difficult to maintain;
– Assets difficult to control.
• Dispersed staff:
– Supervision vs. management;
– Management only long term;
– Supervision must be strong;
– Fast decision taking locally.
• Issue raised by Schmid.
UNIVERSITYOF
Slide No: 19 University of Wollongong, 2012-22-23 BIRMINGHAM

20. Determinants of Railway System Complexity and their Management
Major Structures to cope with Linearity
UNIVERSITYOF
Slide No: 20 University of Wollongong, 2012-22-23 BIRMINGHAM

21. Determinants of Railway System Complexity and their Management
Classic Example of Dispersion Issue
VH
H
M
L
UNIVERSITYOF
Slide No: 21 University of Wollongong, 2012-22-23 BIRMINGHAM

22. UNIVERSITYOF
BIRMINGHAM
Railways are Long and
Thin – they are Dispersed

23. Determinants of Railway System Complexity and their Management
Turn off towards Buxton and Sheffield
1997
1967
2008
2004
1850 to 1895 UNIVERSITYOF
Slide No: 23 University of Wollongong, 2012-22-23 BIRMINGHAM

24. Determinants of Railway System Complexity and their Management
A Difficult Issue: Asset Life Diversity
The extent to which assets of
widely differing ages must
work together to achieve the UNIVERSITYOF
Slide No: 24 purpose of a system (Schmid).
University of Wollongong, 2012-22-23 BIRMINGHAM

25. Determinants of Railway System Complexity and their Management
Asset Life Diversity: 1 day to 200 Years
• Long life railway assets:
– Cuttings, embankments;
– Culverts, bridges, viaducts,
flyovers, dive-unders, tunnels;
– Stations, offices, depots.
VH
• Medium life railway assets:
H – Tracks, rails and signals;
– Locos, carriages, ferries;
M – Wagons, track machines,
• Short lived railway assets:
L – Ticket machines, ticket gates;
– Computers, cars and trucks;
– Staff uniforms and hand-tools.
• Issue raised by Armitage.
UNIVERSITYOF
Slide No: 25 University of Wollongong, 2012-22-23 BIRMINGHAM

26. UNIVERSITYOF
BIRMINGHAM
We are using Assets that
were created between just
5 and 160 Years ago

27. UNIVERSITYOF
BIRMINGHAM
We are still on our
winter journey from
Manchester to Sheffield

28. Determinants of Railway System Complexity and their Management
In Theory, it’s only 6 Mins to Chinley
‘Slow’ Route from New Mills
joins from left, ‘stopper’ waiting
… and ‘stopper’ still waiting
UNIVERSITYOF
Slide No: 28 University of Wollongong, 2012-22-23 BIRMINGHAM

29. Determinants of Railway System Complexity and their Management
Sadly, that’s not what happened on the
stormy Saturday, 8 January 2005
A Level Crossing
Woodsmoor
Station
There are 7500 of UNIVERSITYOF
Slide No: 29 BIRMINGHAM
these in Britain
University of Wollongong, 2012-22-23

30. Determinants of Railway System Complexity and their Management
Determinant: Nature of Interactions
Way in which subsystems and
activities relate to each other
during normal and disturbed UNIVERSITYOF
Slide No: 30
operations of Wollongong, 2012-22-23
University (after Perrow) BIRMINGHAM

31. Determinants of Railway System Complexity and their Management
‘Complex’ Interactions (Perrow)
• Linear Interactions:
– Segregated subsystems;
– Easy substitutions;
– Few feedback loops;
– Single purpose, separate controls;
– Direct information;
– Extensive understanding.
• ‘Complex’ interactions:
– Parts and units not in a production
sequence are close together;
– Unfamiliar or unintended
feedback loops;
– Indirect or inferential information
sources;
– Limited understanding of some
processes.
UNIVERSITYOF
Slide No: 31 University of Wollongong, 2012-22-23 BIRMINGHAM

32. Determinants of Railway System Complexity and their Management
Life is not always easy for railways…
UNIVERSITYOF
Slide No: 32 University of Wollongong, 2012-22-23 BIRMINGHAM

33. Determinants of Railway System Complexity and their Management
Linear vs. ‘Complex’ Interactions
• Railway interactions:
– Train sequences;
– Conflicts at junctions;
– Passenger behaviour;
C – Staff behaviour;
– Trespass and vandalism.
• Maintenance activities:
– Rolling stock and track;
N – Structures and stations.
• System control:
– Management of nodes;
– Staff allocation to duties;
L – Rolling stock schedules.
• Issue raised by Perrow.
UNIVERSITYOF
Slide No: 33 University of Wollongong, 2012-22-23 BIRMINGHAM

34. Determinants of Railway System Complexity and their Management
The Train stops at Woodsmoor at 07:34
Woodsmoor
Sheffield M.
Chinley R2
R1 Chesterfield
UNIVERSITYOF
Slide No: 34 University of Wollongong, 2012-22-23 BIRMINGHAM

35. Determinants of Railway System Complexity and their Management
… and conductor provides information
• We are stuck behind a Class 323 without a pantograph:
– Branch of a tree had broken off and was foul of OHLE;
– First electric train of the morning (in the dark) hit branch;
– Pantograph and branch combine to break droppers;
– Pantograph disintegrates – no power.
• Class 158 is hemmed-in, in both directions:
– Class 323 ahead without power is unable to move;
– Level crossing behind train only strikes-in in one direction;
– The cross-overs of Hazel Grove are beyond the 323.
• Access is near impossible due to housing and gardens.
UNIVERSITYOF
Slide No: 35 University of Wollongong, 2012-22-23 BIRMINGHAM

36. Determinants of Railway System Complexity and their Management
Let’s Look at the Situation in Detail
Rather than here, at Hazel Grove
Failed
Train
Here
Platform at Woodsmoor
UNIVERSITYOF
Slide No: 36 University of Wollongong, 2012-22-23 BIRMINGHAM

37. Determinants of Railway System Complexity and their Management
Many Subsystems and Interfaces
Electrification & Traction &
Vehicle
Power Supplies Braking Structures
Third Parties
Operations Management
Systems
Maintenance System
Bogies
Communi- VCS
Systems
Control
cations &
Axles & Wheels
Signalling
Systems
CIS Rail ATP Rail
Sleepers & Ballast
Station Systems
Substructure System
CIS: Customer Information Systems / VCS: Vehicle Control Systems
UNIVERSITYOF
Slide No: 37 University of Wollongong, 2012-22-23 BIRMINGHAM

38. Determinants of Railway System Complexity and their Management
Interfaces: Strength of System Coupling
Extent to which two components or
activities must be linked to achieve an
UNIVERSITYOF
Slide No: 38 appropriate performance (after Perrow).
University of Wollongong, 2012-22-23 BIRMINGHAM

39. Determinants of Railway System Complexity and their Management
Loose Coupling vs. Tight Coupling
• Loose Coupling (process and technical):
– Processing delays possible;
– Sequence & order can be changed;
– Alternative methods available;
– Slack in resources possible;
– Buffers and redundancies are fortuitously always available.
• Tight Coupling (process and technical):
– Time-dependent behaviour, i.e., delays in processing not possible;
– Invariant sequencing;
– Only one method to achieve goal;
– Little slack available;
– Buffers and redundancies must be designed in as part of the system.
• Issue raised by Perrow.
UNIVERSITYOF
Slide No: 39 University of Wollongong, 2012-22-23 BIRMINGHAM

40. Determinants of Railway System Complexity and their Management
Tightly Coupled Mechanical Systems
• Wheel and rail interface:
– Steel on steel stiffness;
– Motion control by conicity;
– Traction and braking with
small contact area;
– Track held by ballast.
• Pantograph and overhead:
– Overhead relative to rails;
– Low contact force limits wear;
• Switches and crossings:
– Accurate mechanisms needed;
– Must be locked for trains.
UNIVERSITYOF
Slide No: 40 University of Wollongong, 2012-22-23 BIRMINGHAM

41. Determinants of Railway System Complexity and their Management
Railways are Tightly Coupled Systems
• Single degree of freedom of movement of rolling stock
requires infrastructure with variable geometry;
• Limited adhesion requires train control and signalling;
T
• Stiffness of wheel / rail interface requires accurate
infrastructure and high quality maintenance;
• Linear (distributed) nature of the railway infrastructure
propagates failures and is open to environmental
L
influences;
• Need for reliable timetabled operation and good
resource management;
• Interface and interaction management is essential.
UNIVERSITYOF
Slide No: 41 University of Wollongong, 2012-22-23 BIRMINGHAM

42. Determinants of Railway System Complexity and their Management
Charles Perrow 1999
• “[S]ystems are not linear or complex, strictly speaking,
only their interactions are. Even here we must recall
that linear systems have very few complex
interactions, while complex ones have more linear
ones, but complex interactions are still few in number”
• This is often referred to as ‘Interactive Complexity’:
– Do we prefer linear or complex systems?
– Do we prefer loose or tight coupling?
• Unfortunately, some systems have to be interactively
complex and tightly coupled by their nature!
UNIVERSITYOF
Slide No: 42 University of Wollongong, 2012-22-23 BIRMINGHAM

43. Determinants of Railway System Complexity and their Management
System Dimensions and Industries
Interaction / Coupling Chart
Tight Coupling
dams nuclear
power railways >> plant
grids DNA
nuclear
weapons
Linear
Interaction Complex
Interaction
assembly-line
production mining
R&D firms
simple-goal
agencies Loose
Coupling
UNIVERSITYOF
Slide No: 43 Adapted from Perrow,University of Wollongong, 2012-22-23
1999:97 BIRMINGHAM

44. UNIVERSITYOF
BIRMINGHAM
After 20 minutes, Felix
‘jumps ship’, illegally

45. Determinants of Railway System Complexity and their Management
There was another Route to Sheffield…
R3
Sheffield M.
Chinley R2
R1 Chesterfield
UNIVERSITYOF
Slide No: 45 University of Wollongong, 2012-22-23 BIRMINGHAM

46. UNIVERSITYOF
BIRMINGHAM
But the Driver’s Route
Knowledge had Lapsed,
due to the Sickness Leave

47. Determinants of Railway System Complexity and their Management
Consequences of just a few Mistakes
• Consequences for the passengers on the Class 158:
– 07:18 should have arrived in Sheffield at 08:09 but, instead,
reaches South Yorkshire at 10:53;
– Felix arrives in Sheffield at 10:10 on the 09:18 and is in the
University 10 minutes before his lecture;
– Most passengers have had to wait for 150 minutes without
decent information – coming from an airport;
– Passengers bound elsewhere have missed connections;
– Railway has lost a lot of goodwill.
• Network Rail and TOC B share ca. £100,000 cost of
delay penalties – both had made serious mistakes.
UNIVERSITY
Slide No: 47
… but Felix has acquired a really excellent case-study!
University of Wollongong, 2012-22-23 BIRMINGHAM
OF

48. Determinants of Railway System Complexity and their Management
Case Study Lessons I
• Accumulation of minor mistakes and failures can lead
to a collapse in any tightly coupled (transport) system:
– Single degree of freedom of motion of rolling stock:
• Railway requires points;
– Limited adhesion and thus inability to drive by line of sight:
• Railway requires signalling and formal (level) crossings;
– Stiff interface requiring highly performing maintenance:
• Minimise scale of infrastructure to limit cost;
– Linear nature of railway increases management difficulty:
• Reduces ability to control and intervene;
– Operational characteristics require adherence to timetable.
• Organisational structure of system affects outcomes.
UNIVERSITYOF
Slide No: 48 University of Wollongong, 2012-22-23 BIRMINGHAM

49. Determinants of Railway System Complexity and their Management
Case Study Lessons II
• Impact of natural characteristics of rail mode of
transport:
– Route learning failure (driver should not have resumed work
without it);
– Train design failure (two pantographs would have
minimised knock-on delays);
– Control and supervision failure (better management and
early intervention);
– Infrastructure and signalling inadequacies (no capability for
reversal or overtaking of failed train);
– (Environmental) unpredictability not factored in.
• The railway does not forgive mistakes!
UNIVERSITYOF
Slide No: 49 University of Wollongong, 2012-22-23 BIRMINGHAM

50. Determinants of Railway System Complexity and their Management
Determinant: Railway Variability
Extent to which tasks must
depart from a constantly
recurring simple pattern UNIVERSITYOF
Slide No: 50 (McKechnie).
University of Wollongong, 2012-22-23 BIRMINGHAM

51. Determinants of Railway System Complexity and their Management
Not enough Variability to cause Trouble
Photograph Courtesy Keeping Track Image Library
UNIVERSITYOF
Slide No: 51 University of Wollongong, 2012-22-23 BIRMINGHAM

52. Determinants of Railway System Complexity and their Management
Too much Variability to cause Trouble?
Photographs Courtesy North Sout Railway
UNIVERSITYOF
Slide No: 52 University of Wollongong, 2012-22-23 BIRMINGHAM

53. Determinants of Railway System Complexity and their Management
External and Internal Variability
• External variability:
– Operational impact of weather;
– Demand variation;
– Economic cycle impact;
VH – Stakeholder vacillation;
– Subsidy regime variation;
H – Impact of connecting services;
– Third party behaviour.
M • Internal variability:
– Variable passenger behaviour;
L – Variable staff performance;
– Variable wheel-rail adhesion;
– Spontaneous system failures;
• Issue raised by McKechnie.
UNIVERSITYOF
Slide No: 53 University of Wollongong, 2012-22-23 BIRMINGHAM

54. UNIVERSITYOF
BIRMINGHAM
Railways Operate in a
Highly Variable
Environment
Physically, Operationally and
Organisationally

55. Determinants of Railway System Complexity and their Management
And here it is: The Butterfly of Railway Complexity
Le Papillon de la Complexité
Ferroviaire
Der Komplexitätsschmetterling UNIVERSITYOF
Slide No: 55 University of Wollongong, 2012-22-23
des Systems Bahn BIRMINGHAM

56. UNIVERSITYOF
BIRMINGHAM
A Slightly Different
Perspective

57. Determinants of Railway System Complexity and their Management
Regulations and Standards Issue
• Tools to control complexity?
• Intricate legal framework:
!
RS
E – Interoperability regulations;
EA
– Technical Specifications for
Interoperability;
5Y
VH
– Road traffic regulations;
R
– National health and safety law;
TE
H
– European rail safety law.
AF
M • Intricate standards system:
TE
– CEN Standards;
DA
L – UIC ‘standards’;
OF
– National regulations;
– Internal standards.
T
• Issue raised by Qurashi. OU
UNIVERSITYOF
Slide No: 57 University of Wollongong, 2012-22-23 BIRMINGHAM

58. Determinants of Railway System Complexity and their Management
Five Dimensions of Complexity
Variability Ve
ry ?
McKechnie S hor
t
gh
y Hi Sh
ort
r
Ve
Water gh Me
Hi diu
Rail m
Nuclear m
diu
Very High
Me Hig
Medium
h
w
High
Low
Lo
Regulations Diversity
Various
& Standards
Low
High
Very High
Medium
Excessive!
Qureshi
Low Low
um
Medi
Me
diu
m h
Hig
Hig
h h
ry Hig
Ve
Ver
yH
igh
Dispersion Interdependence
Schmid UNIVERSITYOF
Perrow and McKechnie
Slide No: 58 University of Wollongong, 2012-22-23 BIRMINGHAM

59. Determinants of Railway System Complexity and their Management
What Core Issue have we forgotten?
b le
s ha
P eri
RY
e VE
ar
ay
ilw
Ra
f the
c ts o
du
Pr o
The
UNIVERSITYOF
Slide No: 59 University of Wollongong, 2012-22-23 BIRMINGHAM

60. Six Dimensions of Complexity Determinants of Railway System Complexity and their Management
Variability Ve
ry Product Life
McKechnie Sh
ort Harris
gh
y Hi Sh
ort
r
Ve
Water gh Me
Hi di um
Rail
Nuclear m
diu
Very High
Me Hig
Medium
h
w
High
Low
Lo
Regulations Diversity
Various
& Standards
Low
High
Very High
Medium
Excessive!
Qurashi
Low Low
um
Medi
Me
diu
m h
Hig
Hig
h h
ry Hig
Ve
Ver
yH
igh
Dispersion Interdependence
Schmid UNIVERSITYOF
Perrow and McKechnie
Slide No: 60 University of Wollongong, 2012-22-23 BIRMINGHAM

61. Determinants of Railway System Complexity and their Management
Summary of Complexity Determinants
• Level of Variability:
– Affects extent to which tasks must depart from simple recurring patterns.
• Level of Diversity (heterogeneity):
– Relates to number of distinct and different sub-activities that must be
performed within an integrated system of tasks;
– Affects extent to which assets and processes exist beyond normal
planning and management horizons.
• Level of Interdependence (tight coupling / intensive interactions):
– Extent to which performance of a system, as a whole, is reliant on and
facilitated by exchanges of information to co-ordinate individual tasks.
• Level of Dispersion and Linearity:
– Extent to which assets, resources and people contributing to correct
operation of system are distributed over a large area / along corridors.
• Regulations and Standards:
– Extent to which activities are regulated by governments and other
bodies. UNIVERSITYOF
Slide No: 61 University of Wollongong, 2012-22-23 BIRMINGHAM

62. Determinants of Railway System Complexity and their Management
Physical Characteristics of Rail Mode
Characteristic → Motion restricted to single Low coefficient of friction Stiff interface between wheels Distributed linear infra-
↓ Aspects degree of freedom along track between wheels and rails and rails structure subsystem
Strengths G • No steering required; • Low rolling resistance; • Low energy dissipation; • Product reaches customer;
• Predictable motion; • Low rolling surfaces wear; • High tonnages / period; • Production process controll-
• Narrow swept path; • Efficient propulsion; • Low forces in track bed; able throughout system;
• Linked consists (trains); • High speed operation; • Predictable motion; • External events rarely affect
• High standard of safety. • Energy efficiency. • Smooth operations; all of system;
• Potentially long track life. • Part opening of new
systems.
E • Track-based power supply. • Energy recovery potential. • Low wheel-rail damping. • Multiple feeder options.
Weaknesses G • Guidance function cost; • Limited braking rate; • Stiff rolling interface; • Environmental impact
• High route blockage risk; • Low acceleration rate; • Low inherent damping; affects linear strips of
• Low network flexibility; • Seasonal adhesion variation; • Noise & vibration issues terrain;
• Complex route changes; • Line of sight inadequate; • Cost of track & structures; • Remote management of
• No collision avoidance. • Low rolling surface wear. • Cost of inspection. local problems difficult.
E • Complex electrification; • Risk of slip and slide; • High impact environment • Voltage drop along route;
• Limited design options. • Torque control required. for traction drives. • Many supply points needed.
Technical G • Variable geometry elements; • Signalling system; • Load rack design; • Provision of redundancy;
requirements • Train position detection; • Adhesion control; • Testing & inspection; • Protective features (tunnels,
• Locking of route elements; • Artificial wear required; • Accurate maintenance; galleries, fences etc.).
• Junctions & stations. • Regular maintenance. • Regular maintenance.
Operational G • Timetabling & planning; • Path allocation to trains; • Strong procedures. • Scheduling of services;
requirements • Strict rulebook for all staff. • Stringent safety rules. • Several layers of control.
Management G • Rigorous selection of staff; • Simulation of individual • Maintenance management; • Delegated authority;
tools • Modelling of train services. train behaviour. • Technical understanding. • Strong supervision.
Training G • Responsibility; • Environmental awareness; • Strong engineering skills; • Rule based behaviour;
• Staff competence. • Safety ethos. • Safety ethos. • Adaptive behaviour.
UNIVERSITYOF
Slide No: 62 University of Wollongong, 2012-22-23 BIRMINGHAM

63. Determinants of Railway System Complexity and their Management
Railways have Strong Competitors
Constraints / Controls
Timetable, Management Systems
Road Transport
Demand Transport
Product
Profit?
Transport Rail Transport Transport Goods
Demand Demand & People by Rail Service
Quality
Air Transport
Demand
Mechanisms / Processors
Waterways People, Rolling Stock, Infrastructure,
Demand Power, Supplies, Finance
UNIVERSITYOF
Slide No: 63 University of Wollongong, 2012-22-23 BIRMINGHAM

64. Determinants of Railway System Complexity and their Management
How can we Design better Systems?
• Understand better the purpose of systems:
– Identify stakeholder requirements;
– Make stakeholder requirements measurable.
• Understand better the functions that satisfy purpose:
– Identify system functionality and architecture;
– Identify necessary subsystems and their functions.
• Understand better the interfaces between subsystems:
– Identify links and relationships between functions.
• Understand fully the interactions between subsystems:
– Define limits to the interactions;
– Monitor the interactions.
UNIVERSITYOF
Slide No: 64 University of Wollongong, 2012-22-23 BIRMINGHAM

65. UNIVERSITYOF
BIRMINGHAM
It’s a Complex System:
Let’s apply (Railway)
Systems Engineering

66. Determinants of Railway System Complexity and their Management
Rail Systems Engineering & Integration
• Railway Systems Engineering and Integration (RSEI) is
concerned with:
– Managing the people, resources and processes required to conceive,
design, build, operate, maintain, renew, close and decommission
railways of all types;
– Respecting the limits and constraints imposed by the natural physical,
organisational and operational characteristics of the rail mode, in an
effective and efficient manner;
– Satisfying the system’s stakeholders and environmental concerns.
• RSEI is not just about technologies, components, interfaces,
know-how and processes;
• RSEI is about developing people to carry out their tasks better
and more effectively, while respecting the constraints of a
highly complex technical and organisational system.
• ‘Integration’ goes beyond systems engineering…
UNIVERSITYOF
Slide No: 66 University of Wollongong, 2012-22-23 BIRMINGHAM

67. Determinants of Railway System Complexity and their Management
History of Systems Engineering
© Brian Halliday, Network Rail
1950 1960 1970 1980 1990 2000
Products Software Projects
UNIVERSITYOF
Slide No: 67 University of Wollongong, 2012-22-23 BIRMINGHAM

68. Determinants of Railway System Complexity and their Management
Systems Engineering Responsibilities
• Requirements Management:
– Requirements elicitation and requirements management;
– Definition of system and subsystem specifications.
• Performance and Technical Risk:
– Modelling of operational performance;
– Evaluation of robustness of technical options.
• Cost and Capability:
– Development of optimised system options;
– Human, equipment and operations integration.
• Quality Systems Design:
– Creation of quality management systems.
• Configuration Management:
– Provision and control of asset information;
– Development of configuration management systems.
UNIVERSITYOF
Slide No: 68 University of Wollongong, 2012-22-23 BIRMINGHAM

69. Determinants of Railway System Complexity and their Management
Typical Systems Engineering VEE Process
Conceive and Concept of Operation Need for a New
Select Solution Function / System
Capture Stakeholder Commission into
Requirements Service Operation
Develop System Define Validation Plan Test Whole System
Specification Functionality
De
fin
n
tio
it
ion
da
Design and Develop Define Test Plans Integrate System
,D
li
Va
System Architecture Components
eco
&
mp
ing
osi t
st
Te
i on
Develop Components Test Test Components
n,
&
tio
and Subsystems Plans and Subsystems
Ve
ra
ri
teg
fi c
In
a ti
on
Source or Produce
System Components
UNIVERSITYOF
Slide No: 69 Project Time Line
University of Wollongong, 2012-22-23 BIRMINGHAM

70. Determinants of Railway System Complexity and their Management
Typical Systems Engineering VEE Process
Conceive and Concept of Operation Requirement
Select Solution Satisfied
Capture Stakeholder Commission into
Requirements Service Operation
Develop System Define Validation Plan Test Whole System
Specification Functionality
De
fin
n
tio
it
ion
da
Design and Develop Define Test Plans Integrate System
,D
li
Va
System Architecture Components
eco
&
mp
ing
osi t
st
Te
i on
Develop Components Test Test Components
n,
&
tio
and Subsystems Plans and Subsystems
Ve
ra
ri
teg
fi c
In
a ti
on
Source or Produce
System Components
UNIVERSITYOF
Slide No: 70 Project Time Line
University of Wollongong, 2012-22-23 BIRMINGHAM

71. Determinants of Railway System Complexity and their Management
Purpose of VEE Process in SE&I
• Ensures structured approach to projects;
• Assists robust requirements capture and maintenance;
• Ensures clear staging of projects:
– Allows establishment of stage gates;
– Allows monitoring of time line;
– Encourages closing out of issues;
– Encourages management of prevarication.
• Encourages thinking ahead to later stages;
• Ensures robust verification and validation;
• Assists robust configuration management:
– Reduction of modification effort and rework.
UNIVERSITYOF
Slide No: 71 University of Wollongong, 2012-22-23 BIRMINGHAM

72. Determinants of Railway System Complexity and their Management
Why do projects fail?
Other
23% Lack of User Input
13%
Incomplete
Require- 12%
Technology ments
Changing
11% 48%
12%
Unrealistic Time Unrealistic
4% 6%
Inadequate Unclear
Resources 5%
6% Lack of Executive
Support
8% Standish CHAOS report, 1995, http://standishgroup.com/visitor/chaos.htm
UNIVERSITYOF
Slide No: 72 University of Wollongong, 2012-22-23 BIRMINGHAM

73. Determinants of Railway System Complexity and their Management
What goes wrong?
50%
40%
30%
20%
10%
0%
Incorrect facts Omissions Inconsistency Ambiguity Misallocation
Leffingwell, http://www.rational.com/media/whitepapers/roi1.pdf
UNIVERSITYOF
Slide No: 73 University of Wollongong, 2012-22-23 BIRMINGHAM

74. Determinants of Railway System Complexity and their Management
How much does it cost to fix?
250
200
200
Relative Cost to Fix
150
100
50
50
20
1 5 10
0
Requirements Design Build Test Commissioning Operation
Lifecycle Phase when Error Discovered
Leffingwell, http://www.rational.com/media/whitepapers/roi1.pdf UNIVERSITYOF
Slide No: 74 University of Wollongong, 2012-22-23 BIRMINGHAM

75. Determinants of Railway System Complexity and their Management
Conclusion
• Natural characteristics of rail mode are constraining;
• Railways are fundamentally different from other
modes of transport and service industries;
• Rail mode is inherently complex:
– Technologically, organisationally and operationally;
– Project delivery and safety management.
• Stakeholders often have conflicting requirements and
contradictory agendas;
• VEE project life-cycle can be helpful in major projects
but requires strong PM who imposes stop criteria!
UNIVERSITYOF
Slide No: 75 University of Wollongong, 2012-22-23 BIRMINGHAM

76. UNIVERSITYOF
BIRMINGHAM
Railways perform well
when they observe a robust
and efficient timetable