In the second article of our series on the CulturalRoad pillars of mobility equity, we examine how network optimisation plays a crucial role in supporting and enhancing public transport. In CulturalRoad, the Network Optimisation pillar aims to develop guidelines for the future integration of Connected, Cooperative and Automated Mobility (CCAM) services within multimodal transport networks, encompassing both public and private modes. This integration seeks to enhance the overall efficiency, coverage, and sustainability of the transport system.
The final objective is to calculate and interpret Key Performance Indicators (KPIs) to assess the impact of different implementation strategies across the project’s demonstration sites. Achieving this requires not only that CCAM solutions are adequately integrated within the public transport eco-system, but also that the different elements of the public transport systems are jointly optimised with a specific focus on equity, and diversity. The pillar therefore investigates optimisation tools that allows to better manage mixed fleets composed of different vehicles (such as CCAM and conventional human-driven vehicles) and various operational settings (including frequency, scheduling, on-demand). The goal is to optimise not only operational costs but also transport equity. This will be done through the integrated optimisation of schedule-based and frequency-based public transport lines, and simulation-based optimisation for large scale scenarios.
What is network optimisation?
Network optimisation refers to the process of improving how different elements of a transport system, such as routes, vehicles, and schedules, work together to deliver efficient, sustainable, and accessible mobility. It involves strategic decisions about service frequency, vehicle allocation, and route planning to ensure that transport services meet the needs of all users.
In public transport, frequency-based services operate at regular intervals without a strict timetable, allowing passengers to arrive at stops and expect a vehicle shortly. Metro systems or bus lines conventionally follow this model and are well suited for high-demand corridors. In contrast, schedule-based services follow fixed timetables and are often used in areas with lower demand or longer travel distances – for example, regional trains (as well as long distance trains), as well as bus lines operating at a low frequency (e.g., a few times a day) between different urban centres.
To investigate the optimisation of joint fleets of conventional CCAM services in the context of public transport, the Network Optimisation pillar will focus on the optimisation of timetables, capacity, allocation of seats for vulnerable users, the impact of technological constraints in terms of operations, and the possibility to replicate the dynamics of large scale scenarios.
How can CCAM support public transport?
CCAM solutions can enhance both service types by enabling flexible, demand-responsive options, and supporting more efficient coordination across mixed fleets, including both automated and human-driven vehicles. Crucially, CCAM technologies can help promote transport equity by better serving diverse user needs, such as people with reduced mobility or users living in underserved areas, through optimised coverage, adaptive service models, and integrated planning across the transport network. Examples include using robotaxis for on-demand services in low density areas, increased frequency of night services, and better integration of frequency and scheduled based services.
“The integration of public transport and CCAM services is of key importance. If we simply replace private cars with CCAM solutions, even in the case of robotaxis, we are going to miss out on the opportunity of CCAM. Autonomous vehicles alone cannot solve the congestion and (space) occupancy problems. Public transport should be the backbone of urban commuting, and CCAM should contribute to enhance it, not compete with it. CCAM stand for cooperative, connected, and automated mobility. Public transport fleets are the most effective way to fully take advantage of their cooperative nature, contributing to a more efficient, equitable, and sustainable mobility paradigm.” Guido Cantelmo, Assistant Professor in Transport Modelling, Technical University of Denmark (DTU)
Initial results
Led by the Technical University of Denmark (DTU), the Network Optimisation pillar is already producing insightful results. We expect to see that optimisation of CCAM services with and without explicit consideration of vulnerable user may cause inequalities, possibly worsen the current situation for vulnerable users instead of improving it.
Initial results suggest that optimising public transport systems for a single, homogeneous user group leads to suboptimal outcomes when the system is evaluated under heterogeneous demand. Specifically, applying a solution tailored to a fully uniform user base in a mixed-user scenario results in a noticeable increase in overall system costs. These additional costs are not distributed evenly and tend to disproportionately affect vulnerable users.
In contrast, when the diverse needs of different user groups are considered during the optimisation process, the resulting transport system performs equitably across the entire population. Notably, this inclusive approach does not improve service for vulnerable users by reducing the quality of service for others. Instead, it enhances the level of service for all commuters equally, demonstrating that planning for diversity does not inherently compromise efficiency.
These findings suggest that inclusive, user-sensitive transit optimisation can produce more robust and equitable outcomes, while ignoring user heterogeneity may lead to inefficiencies and increased social costs.
A detailed account has been submitted to the EWGT 2025 Conference panel. If accepted, the paper entitled “Network Optimization for Mixed-Use Transit: Accounting for Vulnerable Users”, will be presented by DTU at EWGT 2025 in Edinburgh (1-3 September). Stay tuned for the publication on our website!
