System Dynamics at the European Transport Conference

The European Transport Conference (ETC), now in its 48^th year, was due to be held in Milan, Italy, but was held virtually due to the COVID restrictions. Being on the “System Dynamics” programme committee, it was a valuable experience to be involved in the discussions about how the transitions online should take place. In all, and to much praise to the efforts of the organisers, the conference was a success and went smoothly from the point of view of a participant.

The SD theme is new to ETC (in its second year) and supported by the System Dynamics Society Special Interest Group in Transportation. The concept of ‘system dynamics’ as a specific methodology is yet to be fully understood by a wider audience, so having the opportunity to engage with the transport community at such a large event is one much appreciated.

Urban Mobility, Future Mobility and Decarbonisation were the main themes through which the SD concept was investigated at the conference. Thanks to the experiences, studies, and research of the speakers, we had the chance to discover how the SD approach is developing all around Europe. Seven presentations were given over three sessions, showcasing a wide range of applications of SD to transport research and practise.

Case Studies

The application of SD to a feasibility study of a tramway in the city of Bozen – presented by NET Engineering –  was the opportunity to better understand the advantages of using SD for complex transport projects that require economic, social and environmental assessment, developing a Causal Loop Diagram that incorporates the key concerns of stakeholders. Furthermore, the developed Causal Loop Diagram allowed engineers and city managers to have a common understanding of long-term impacts of the new tramway on the city dynamics, useful for further considerations on urban mobility policies.

ETC 2020 – a different conference experience!

On this topic, the University of Leeds is developing an SD model of tradable transport permit schemes. There are complex legal and ethical concepts to be considered within a governance framework for this innovative new form of mobility policy, though previous studies have argued that it has the potential to be more efficient and equitable than comparative road pricing schemes. SD is an ideal approach for capturing key variables and relationships within such a scheme and run scenarios to aid the understanding of actor behaviours and under what conditions policy targets can be realised.

The development of tools is at the centre of the studies on SD. WSP presented at the ETC a SD tool that captures the uptake dynamics and interactions between of electric, automated, and shared vehicles. This was developed with a user-friendly interface for use by policy makers. The University of Natural Sciences (BOKU), Austria, is developing a tool to overcoming implementation barriers and is using a method of the qualitative analysis of interviews (carried out online due to COVID restrictions) regarding integrated mobility solutions, which has been used to develop a Causal Loop Diagram. The University is also working on an educational game that engages children (aged 12-13 years) with transport policy understanding but also introduces them to systems theory.

Concerning the use of SD for “Decarbonisation”, EC-JRC has developed the Powertrain Technology Transition Market Agent Model (PTTMAM, an extensive SD simulation model of light duty vehicle technology development in Europe). At ETC, a study was presented on the impact of supply side mechanisms on EU average new car CO₂ emissions, concluding that policy makers need to carefully consider how specific targets may influence manufacturer behaviours. Complementing this, KTH presented on a SD approach to the impact of the introduction of driverless and electric trucks on meeting 2030 Swedish decarbonisation targets. Uncertainty analyses of six scenarios were considered and can be used to guide decision-makers.

Discussions and insights

All three sessions led to interesting discussions around the engagement of both the public and policy makers with SD as tools for simple simulation and visualisation. In particular, it was raised that the separate quantitative (stock-flow model) and qualitative (causal loop diagram) elements have benefits and applications for different stakeholders, depending on their background needs. As an SD practitioner, there is a responsibility to understand these needs at the start of the engagement process in order to develop an effective approach to ensuring that the stakeholder is provided with the correct level of information required for their input and requirements. For instance, does the stakeholder need to know the underlying theory of SD in order to engage or would this be unnecessary, possibly occluding the understanding of purpose or overcomplicate their involvement? On the other hand, they may profit from seeing quantification from the start in order to grasp the issues at hand. A simple front-end visualisation may be adequate (and appropriate) for some, who do not require deep understanding of the internal system complexities, rather some idea of cause (an input they make) and effect (the output they see). For those who require deeper understanding or engagement, this may be only on a theoretical level, so quantification is not an appropriate way to engage.

Building on this, in reflection to the current greatest global challenge, another key discussion point was the restrictions that COVID have imposed in collecting data, feedback and engagement in model development. A number of presenters found that their planned methodologies for data gathering and model building – requiring face-to-face contact – had to be moved online, introducing barriers to communication. Furthermore, the direct impact that COVID itself is having on transportation systems across the world could effect any transport related research being carried out at present. Although some interviews may be possible on video conferencing with little loss, it is important to have the correct tools for some engagement, which may require creating, discussing and annotating diagrams for example. Perhaps more difficult is the key SD development technique of Group Model Building. Video conferencing again may allow a number people to meet, but does not provide the same opportunity for a number of people to consider complex diagrams and clearly and constructively discuss potential conflicts in opinions. Innovative approaches to overcome these barriers must not only focus on maximising the communication between different stakeholders and avoiding a poorer quality of data, but also consider the opportunities that may be offered. Development of online methodologies may open up SD model engagement to a wider audience than has been possible before and create new insights and approaches for more holistic model building.