Transfers: How to Turn Individual Lines into a True Public Transport Network

2. A Collection of Lines Can Operate Without Becoming a Network

Imagine a city with twenty perfectly operational bus routes. Each has its own alignment, stops, timetable and vehicles. Every route may perform well individually while providing relatively limited overall connectivity if there is little coordination between them. In that case, we have twenty services sharing the same territory, but not necessarily a network in the functional sense.

The difference becomes apparent when we consider not only the destinations that can be reached directly from a stop, but all the destinations that can reasonably be reached by combining services. A stop served by one route may provide direct access to ten relevant destinations. If that route connects efficiently with three others, the number of accessible opportunities may multiply. The value of each route then depends not only on its own alignment, but also on its position within the wider system.



This is precisely why some of the technical documents reviewed do not limit their assessment to individual route coverage and demand, but also include total network coverage with transfers among their indicators. This concept captures a particularly important idea: the actual reach of a network does not end where the route we boarded terminates.

3. Indirect Connectivity Can Greatly Increase the Value of a Network

Suppose a particular neighbourhood has only one bus route connecting it directly to the city centre. If we assess only direct connections, we may conclude that the accessibility provided is relatively limited. However, if that route reaches an interchange where passengers can connect to services serving a university, a hospital, a railway station and several employment centres, the actual range of travel opportunities becomes considerably greater.

This network effect is fundamental because it allows a limited number of routes to provide many different journey combinations. Reproducing all those combinations through direct services would normally require substantially more routes and resources. Connectivity through transfers, by contrast, makes it possible to concentrate service provision along selected corridors and subsequently distribute passengers towards different destinations.



However, this effect only materialises if the connection is genuinely usable. A theoretically possible connection between two routes requiring a 35-minute wait hardly provides the same benefit as one involving a five-minute wait. Similarly, two routes operating through the same area do not necessarily provide a good connection if their stops are several hundred metres apart or passengers must cross a major infrastructure barrier. Connectivity should therefore be measured not only by the existence of a connection between services, but also by its quality.

4. The Problem Is Not Changing Vehicles, but Everything That Happens During the Transfer

The expression "transfer penalty" may suggest that there is a single cost associated with changing routes. In reality, passengers experience several penalties simultaneously. There is the physical time required to move between vehicles, the waiting time before the next service, any additional fare, uncertainty about the connection, and the additional effort required to navigate and understand the system.

Consequently, two five-minute transfers can be completely different. One may take place on the same platform, under shelter, with real-time information and a high-frequency connecting service. The other may require crossing a major road, locating a poorly signposted stop, waiting without knowing when the next vehicle will arrive, and using a different ticket. The elapsed time may be similar, but the passenger experience is not.



This interpretation is consistent with the way contemporary studies increasingly assess service quality. Transfer time appears alongside frequency, service regularity, commercial speed, stop location and stop infrastructure as a specific diagnostic variable. Transfers should therefore not be analysed as isolated events, but as a combination of different components that can be designed and improved.

5. Total Journey Time Matters More Than the Number of Routes Used

When comparing two travel alternatives, there is a tendency to automatically favour the one that does not require a transfer. However, from a transport planning perspective, the comparison should consider the entire journey. A direct connection may be slow and infrequent, while an alternative involving a transfer may offer a shorter overall journey time.

Consider a direct service operating every 30 minutes and taking 45 minutes to reach the destination, compared with an alternative consisting of two routes operating every 10 minutes, with a total journey time of 35 minutes including the transfer. The second option involves changing vehicles but may be objectively more competitive. Naturally, reliability, comfort and ease of interchange must also be considered, but this example illustrates why the number of vehicles used should not, by itself, determine network design.



In a previous article, we discussed the importance of assessing the complete journey by considering access, waiting, in-vehicle and final access times. In a network that relies partly on transfers, this approach becomes even more important, because a seemingly small improvement in transfer time can simultaneously improve the competitiveness of numerous journeys.

6. Service Frequency Largely Determines Whether a Transfer Is Acceptable

The frequency of connecting services is probably one of the most important variables. When two routes operate every five or ten minutes, the need to consult timetables decreases considerably. Passengers can transfer relatively spontaneously because they know the next vehicle will arrive shortly. When headways increase to 30, 45 or 60 minutes, the situation changes completely.

In these cases, a small difference between timetables can turn a good connection into a very poor one. If a bus consistently arrives five minutes after the connecting service has departed, passengers may have to wait almost an entire service cycle. A relatively minor timetable adjustment can therefore generate substantial benefits without adding vehicle-kilometres or additional vehicles.



It is particularly significant that, when considering the integration of urban and interurban services, the technical documents reviewed refer not only to the provision of interchange facilities, but also to timetable coordination and fare integration. Infrastructure makes the physical transfer possible; the timetable determines whether it is reasonable in terms of time.

7. The Lower the Frequency, the More Important Timetable Coordination Becomes

In a very high-frequency urban network, synchronising every arrival with every departure may be unnecessary. Another service will arrive within a few minutes, and strict coordination could even complicate operations. However, in metropolitan, interurban or rural services operating at longer intervals, coordination becomes considerably more important.

A connection between two hourly services can work exceptionally well when the transfer requires only a few minutes, but it can become practically unusable if the waiting time approaches one hour. The frequency of each individual route has not changed; what changes is how effectively they operate together.



This highlights an important distinction between individual route frequency and network-level service frequency. A relatively infrequent service can become substantially more useful when it provides efficient connections with other services. Timetable planning then ceases to be an exclusively route-specific task and becomes a tool for designing the entire network.

8. A Connection Only Works If Passengers Can Rely on It

Timetable coordination immediately introduces another challenge: reliability. There is little benefit in designing a theoretical three-minute connection if the first service frequently arrives five minutes late. The tighter the connection, the more vulnerable it may become to small operational variations.

This requires frequency, regularity and transfer time to be considered together. In some cases, allowing a certain buffer between services will be appropriate. In others, sufficiently high frequency will make it unnecessary to hold the connecting vehicle. The appropriate approach will depend on the type of network, the number of passengers affected and the ability to manage operations.



This relationship helps explain why public planning documents include service regularity and transfer time within the same group of quality indicators. An interconnected network requires reasonably predictable services, because irregularity on one route does not affect only its direct passengers: its effects can extend to all journeys that depend on that route for a connection.

9. Interchange Points Should Be Located Where Travel Flows Actually Converge

Not every location where several routes coincide should automatically become an interchange. The location of interchange points should reflect travel patterns and the functional structure of the network. The most important nodes will be those where significant passenger flows converge and where transferring can substantially increase the number of accessible destinations.

It is therefore useful to analyse not only where transfers currently take place, but also their volumes. Among the parameters that some public authorities require to be studied are precisely the locations and volumes of transfers. This information helps identify which connections already function in practice and which could become important following a network redesign.



Furthermore, the introduction of a high-capacity system can completely change these patterns. A new tram, BRT or railway line can turn certain stations into strategic nodes and justify reorganising bus services around them. Indeed, the studies reviewed explicitly propose assessing the impact of new tram lines and intermodal stations on existing urban bus routes.

10. Physical Transfer Distance Matters as Much as the Intersection of Routes

On a schematic map, two services may appear perfectly connected because their routes intersect at the same point. In reality, one stop may be located on one side of a roundabout and the other 250 metres away, beyond several signalised crossings. The connection exists on the map, but it may be unattractive to passengers.

Interchange design should therefore assess actual pedestrian routes, walking distances, crossings, gradients, pavement continuity and how easily passengers can find the connecting service. The high-capacity infrastructure studies reviewed explicitly incorporate connections between stations and pedestrian and cycling routes, and require the physical design of stops, intermodal areas and interchange facilities.



This demonstrates that transfers cannot be addressed through route planning alone. There is also an urban design and architectural dimension. A difference of just a few metres in stop location can affect the transfer times of thousands of daily passengers and, consequently, the usefulness of the entire network.

11. An Interchange Is Not Necessarily a Large Station

When discussing intermodality, we often imagine major railway stations or metropolitan transport terminals. However, a network contains interchange points of very different scales. There may be a major hub connecting rail, urban buses and interurban services, but also a simple connection between two bus routes along an avenue, or a stop where a feeder service from a peripheral area connects with a trunk route.

This hierarchy should be explicitly recognised. Technical documents concerning BRT systems even distinguish between basic stops, transfer stops and stations, requiring the identification of their main characteristics and potential users. Not every node requires the same infrastructure, but each should be designed according to its role within the network.



The objective is not necessarily to construct large buildings. In some cases, moving two stops closer together, improving a pedestrian crossing, providing coordinated passenger information and ensuring adequate weather protection can create a highly effective interchange at relatively modest cost.

12. Universal Accessibility Must Be Maintained Throughout the Transfer

A journey may be accessible on both vehicles but cease to be accessible during the transfer itself. There is little benefit in providing low-floor buses if the connection between stops requires passengers to negotiate stairs, unsuitable gradients, unsafe crossings or spaces that are insufficient for people with reduced mobility.

Station and interchange design must therefore consider platforms, the interface between platforms and vehicles, pedestrian routes and safety conditions. The documents reviewed explicitly require universal accessibility at platforms and the assessment of minimum dimensions, vehicle-platform positioning, lighting, surveillance systems and other safety features.



Accessibility during transfers should also be understood in a broader sense. Older passengers, people travelling with luggage, families with pushchairs and passengers unfamiliar with the network may experience difficulties even when the infrastructure formally complies with physical accessibility regulations. A genuine network should enable all these users to understand and complete their connections with reasonable ease.

13. Safety and Comfort Are Part of Perceived Transfer Time

Five minutes spent waiting in a sheltered, well-lit and busy location may not be perceived in the same way as five minutes at an isolated stop without information and exposed to adverse weather conditions. The objective waiting time is identical, but the passenger experience can be very different.

This dimension also appears in station design requirements, where lighting, CCTV, emergency alert systems, covered waiting areas and the location of stops in well-used public spaces are identified as relevant considerations. These are not secondary issues: they influence passengers' willingness to use a connection and become particularly important at night or on lower-frequency services.



Improving a transfer, therefore, does not exclusively mean reducing the number of minutes involved. It can also mean improving the quality of those minutes. A well-connected network requires interchange nodes where waiting is predictable, safe and sufficiently comfortable.

14. Passenger Information Must Explain the Journey Beyond a Single Route

A network based on connections requires a different passenger information system from one designed for a collection of independent services. If passengers only understand the route they are currently using, it becomes difficult to benefit from the network effect. They need to know which connections are available, where they take place and what alternatives exist if a particular service is missed.

Static information — maps, diagrams and signage — remains important, but dynamic information becomes especially valuable when operational uncertainty exists. Knowing the actual arrival time of a connecting service allows passengers to make informed decisions and reduces some of the uncertainty associated with transferring.



Documents concerning public transport modernisation explicitly link technological development and digitalisation with improvements in service quality and passenger information. In an integrated network, information should also be integrated. It is not enough for each operator to provide accurate information about its own services if passengers must work out the connections between them independently.

15. Fare Integration Can Be as Important as Physical Integration

A connection can be excellent in terms of location and timetable coordination, yet still be penalised through the fare system. If passengers must purchase a new ticket every time they change services, the transfer introduces an additional financial cost that may make the overall journey less competitive.

It is therefore particularly significant that public authorities are addressing effective interchange points, timetable coordination and fare integration together when examining the relationship between urban and interurban services. These three components address three different sources of friction: where passengers transfer, how long they wait and how much the transfer costs.



Fare integration can also change how passengers perceive the system. When a single ticket allows access to several services, passengers can begin to see them as components of the same network, even when they are operated or managed by different organisations or authorities. The fare system then becomes more than a revenue collection mechanism: it also becomes a tool for functional integration.

16. Administrative Boundaries Should Not Become Boundaries for Passengers

One of the greatest challenges arises in metropolitan areas. A journey may begin in one municipality, involve an interurban bus, connect with an urban transport network and end with a railway journey. Each service may fall under different authorities and operators, but from the passenger's perspective, it is a single journey.

The regional planning documents reviewed explicitly recognise this situation and propose paying particular attention to the integration of interurban services with the urban networks of individual municipalities, promoting their joint organisation through interchange points, timetable coordination and fare integration.



This is particularly important because the quality of a journey can be determined by the weakest link in the chain. An excellent interurban service connecting with a poorly coordinated urban network can produce an overall journey that is not competitive. Metropolitan planning must therefore assess connections between networks, rather than focusing exclusively on service performance within individual concessions or administrative areas.

17. High-Capacity Systems Need a Network That Feeds Them

A metro, tram, railway or BRT line should not be assessed exclusively on the basis of people who live or work near its stations. Much of its potential demand may come from passengers reaching it through other services.



The high-capacity system studies reviewed specifically require the assessment of intermodality, transfers, park-and-ride facilities and pedestrian and cycling connections, as well as connectivity for areas not directly located along the new corridor. They also require the analysis of frequencies and travel times both on the new infrastructure and across the wider network.

This perspective changes how infrastructure is evaluated. The question should not simply be how many people live within the catchment area of its stations, but how much territory can efficiently access those stations through the rest of the transport system. A well-designed feeder network can greatly expand the functional catchment area of a high-capacity corridor.

18. But Feeding a Main Line Should Not Mean Imposing Unnecessary Transfers

Reorganising services around high-capacity infrastructure can create a temptation to truncate existing routes and systematically require passengers to use the new corridor. From an operational perspective, this may appear efficient, but whether it benefits passengers depends on the performance of the complete journey.

If the main system offers substantially greater speed, frequency or capacity, the transfer may be clearly beneficial. However, if the time saved on the main corridor is limited and the interchange introduces a significant wait, the new arrangement may make the journey worse than the previous direct service.



Studies should therefore go beyond identifying where services can physically connect. The documentation reviewed requires specific assessments of demand capture, service frequencies and journey times across the network and for individual routes. Transfers should be justified by the overall journey outcome, rather than solely by a network hierarchy rationale.

19. Walking Is Also Part of Intermodality

Public transport journeys normally begin and end on foot. In addition, every transfer involves some pedestrian movement, even when it takes place within a station. The quality of pedestrian routes therefore directly influences the quality of the network.

This relationship is explicitly recognised in network restructuring proposals, where intermodality includes not only transfers between public transport modes but also connections with walking and private cars. High-capacity system studies similarly require the assessment of pedestrian and cycling connections.



This means that improvements to a pavement, crossing or station entrance can also be public transport improvements, even when they do not directly affect any vehicle. If they reduce transfer time or improve the safety of a connection, they improve the network.

20. Cycling Can Expand the Catchment Area of Public Transport Nodes

The same principle can be extended to cycling. A railway or high-capacity public transport station may be accessible on foot within a certain radius, but cycling allows passengers to reach the node from greater distances without using a car.

Contemporary technical documents increasingly incorporate cycling routes explicitly into assessments of intermodality for new infrastructure. This requires consideration of issues such as route continuity, access conditions at stations and the availability of secure bicycle parking.



Once again, the network effect becomes apparent. It is not always necessary to extend a transport route to every origin. For certain travel patterns, improving access to a major node through active modes and subsequently concentrating public transport services along the most important corridors may be more efficient.

21. Private Cars Can Also Form Part of an Intermodal Journey

Although many mobility policies aim to reduce dependence on private cars, certain journeys originating in dispersed areas can be difficult to complete entirely by public transport. In these cases, park-and-ride facilities allow only part of the journey to be made by car, while the section passing through more intensively developed urban areas is completed by public transport.

The studies reviewed explicitly include park-and-ride facilities in the assessment of alternatives and intermodality. Their success depends on the characteristics of the connecting public transport service: frequency, capacity, journey time and the physical convenience of the interchange.



A car park next to a station does not automatically constitute a good intermodal solution. There must be a clear advantage over continuing the journey by car. Once again, the transfer only works when the second part of the journey offers a sufficiently attractive level of service.

22. To Manage Transfers, We Must First Measure Them

One of the longstanding challenges in public transport planning has been accurately identifying where and how transfers occur. Data from individual routes may indicate how many passengers use each service, but they do not always make it possible to reconstruct the complete journey chain.

This is why it is relevant that network studies explicitly require the identification of transfer locations and volumes. With sufficiently structured ticketing systems and other data sources, it is also possible to make progress towards reconstructing passenger journey chains, identifying which routes are used in succession, where transfers take place and how much time elapses between them.



This information is particularly valuable because it makes it possible to distinguish between planned transfers and those that occur spontaneously. A node with a high volume of transfers may require physical improvements even if it was not originally designed as an interchange. Actual passenger demand can therefore reveal the functional structure of the network.

23. Transport Models Must Properly Represent Transfer Penalties

When using a public transport model to analyse a network, the treatment of transfers can significantly influence the results. If changing routes carries almost no penalty within the model, it may assign excessive demand to journeys involving multiple transfers. Conversely, if the penalty is too high, the model may artificially favour direct connections.

It is significant that technical documents require transport models to be sensitive to parameters such as fares, frequency, travel time and transfer time. This means representing not only the routes themselves, but also the actual conditions under which connections take place.



Model validation should subsequently assess whether simulated transfer locations and volumes are reasonably consistent with observed data. The objective is not to make the model avoid or generate transfers, but to reproduce passenger travel choices with sufficient realism.

24. Intermodality Should Be Measured as a Network Performance Indicator

If transfers are considered an essential part of the transport system, specific indicators are needed to evaluate them. Measuring only passenger numbers, vehicle-kilometres operated or route punctuality does not reveal how effectively services work together.

The administrative documents reviewed propose at least two particularly interesting indicators: transfer time and total network coverage achievable through transfers. Where the necessary data are available, these can be complemented by the number of passengers making connections, the nodes where transfers are concentrated, the proportion of journeys requiring a transfer and the evolution of these indicators following network changes.



The key is to move beyond treating intermodality as a purely qualitative characteristic. Describing a station as "intermodal" provides relatively little information. It is far more useful to know how many passengers can connect there, how long the transfer takes, which services are linked, their respective frequencies and how many additional destinations become accessible as a result of those connections.

25. Conclusion: A True Network Enables Passengers to Think About Destinations, Not Routes

Transfers have traditionally been regarded as one of the disadvantages of public transport, and in many cases they are. They interrupt the passenger's journey and introduce additional time, uncertainty and effort. However, assuming that every transfer is necessarily a failure leads to another problem: attempting to provide direct connections between too many origins and destinations through a proliferation of routes that ultimately disperses the available resources.

A genuine network must seek a different balance. Some origin-destination pairs may continue to benefit from direct services, while others will work better through connections between high-frequency routes, feeder services, railways, trams or other high-capacity systems. What matters is that those connections are deliberately designed as part of the journey, rather than simply emerging as an accidental consequence of two services passing through the same location.

Achieving this requires several dimensions to work together. Service frequency must limit waiting times; where frequency is low, timetables must be coordinated. Operations must be sufficiently reliable to ensure that planned connections can be made. Stops and stations must be located and designed to minimise transfer distances and guarantee universal accessibility. Passenger information must explain the entire system rather than individual routes alone. Fares should not unnecessarily penalise changes between services. And urban, interurban and rail networks should be coordinated from the perspective of the passenger's complete journey, even when they belong to different administrative systems.

The studies currently commissioned by public authorities reflect this evolution. They no longer focus exclusively on designing routes: they require analysis of transfer locations and volumes, interchange times, coverage achievable through connections, timetable coordination, fare integration, pedestrian and cycling accessibility, park-and-ride facilities and the physical design of major interchange nodes.

This is probably one of the keys to modern public transport planning. A route allows passengers to travel between certain points. A network allows them to combine services and multiply the opportunities available to them. And the better those transfers are designed, the less important it becomes for passengers to know where one route ends and another begins.

At that point, a transfer is no longer simply an interruption.

It becomes the element that makes the network possible.

Lluis Sanvicens, 2026



Photographs taken in Vienna in 2025. Source: author's own collection.