operating practices. They determine baseline performance. If these are poorly configured, no level of operational excellence can consistently overcome them.
Effective design, therefore, requires two questions: which common causes of variation can be removed or reduced through design choices, and which special causes are likely to challenge the system over its life? Addressing both raises baseline performance while limiting the impact of disruption.
Variation from System Use
Railways are shaped daily by how passengers, freight operators, and staff interact with them. These behaviours introduce further sources of variation. Passenger behaviour directly affects dwell times, which become increasingly variable under high demand. Systems designed to account for this perform better. In Tokyo, active platform management maintains predictable dwell times during peak periods.
On London’ s Elizabeth Line, wide doors and level boarding were adopted specifically to reduce dwell time variability. Where infrastructure does not account for behaviour, variation in dwell time becomes a persistent operational constraint. Even the placement of shelters along a platform may have operational implications when it rains.
Freight services introduce additional variability on mixed-traffic networks. Differences in speed, braking rates, and scheduling flexibility between freight and passenger services reduce usable capacity and complicate recovery from disruption. Even when freight operates to plan, its interaction with faster passenger services introduces structural variation. Operational practices also matter. Variations in driving style, dispatching decisions, and local control actions may be individually small, but when repeated across the network they accumulate into system-level effects. Consistency of practice is therefore critical.
Variation from Unaccounted Change Over Time Railways are long-lived assets operating in a changing environment. When systems fail to adapt, variation increases. Demand growth, technological change, and external conditions all introduce new pressures. Infrastructure designed for lower passenger volumes becomes constrained, resulting in crowding, longer dwell times, and reduced punctuality. On London’ s Jubilee Line, peak demand growth led to significant dwell time increases once design limits were exceeded, reflecting outdated assumptions rather than operational failure.
Technological change can also introduce variation if integration is incomplete. New rolling stock or signalling may outpace supporting infrastructure or operating practices, creating transitional instability. Climate change adds further pressure, exposing assets to conditions beyond historic design tolerances. Heat, flooding, and landslips are now recurring challenges that must be treated as common causes of variation.
Implications for Service Delivery
Variation manifests directly in service delivery. Reliability is the most visible casualty. Railways are tightly coupled systems in which small deviations propagate quickly. Persistent punctuality challenges are often linked to systemic bottlenecks rather than isolated failures.
Capacity is also reduced. High variability forces additional margins into timetables, lowering the number of services that can be operated reliably. Passenger experience deteriorates accordingly, as inconsistent journey times undermine trust and suppress demand. Costs rise as additional rolling stock, staffing, and contingency measures become permanent features of operation.
Addressing Common Cause Variation
W. Edwards Deming’ s insight that most performance problems arise from common causes is highly applicable to railways. Sustainable improvement depends on systemic change rather than reactive intervention.
Projects such as the Elizabeth Line were conceived to remove long-standing structural constraints, raising baseline performance. The Japanese Shinkansen demonstrates how infrastructure design, timetable structure, rolling stock, and operational discipline combine to minimise variation and deliver exceptional reliability.
Continuous adaptation is equally important. Regular timetable revision, demand monitoring, and operational learning prevent small mismatches from becoming embedded. Finally, resilience must be designed into the system. Recovery margins, alternative routings, and operational flexibility ensure that inevitable variation does not escalate into widespread disruption.
Understanding and addressing common causes of variation is therefore central to delivering a reliable, efficient, and resilient railway. www. ciro. org
7