Simulating the transition of transport Infrastructures Toward smart and sustainable Cities – SwITCh

In order to achieve the set objective, the SwITCh team proposed an approach combining different methods which, when used in synergy, aimed to enrich the discussions on mobility among the various stakeholders.

First, the SwITCh team proposed a cross-analysis of the views of scientists and experts (mobility operators, technical services, etc.) on mobility in the future. To this end, a series of interviews and a literature review were conducted. Textual analysis methods were used to carry out this analysis.

Two methods were used to construct prospective scenarios: semi-structured interviews and focus group workshops. A method was proposed to conduct these interviews and workshops and to convert the narrative accounts resulting from these approaches into a set of evaluated indicators. In addition, a method was developed to construct a model for evaluating urban mobility. This method is based on an analysis of the literature and data (existing models and available data). The model to be constructed had to be holistic and usable by all types of actors, whether or not they were experts in urban mobility.

In order to convert prospective scenarios in text form into their evaluation by the evaluation model, four methods were used: 1) a direct interpretation of the textual elements, 2) a statistical analysis based on current and past data from French metropolitan areas, 3) expert rules, and 4) Agent simulation. Agent models are particularly well suited to simulating socio-technical systems. By allowing spatial (and even geographical) and temporal dimensions to be explicitly taken into account, they are relevant for addressing urban mobility issues. The challenge for the SwITCh models was to be able to simulate prospective scenarios in order to evaluate them against the various criteria of the evaluation model.

Finally, in order to put the knowledge developed within the SwITCh project into action, we used “serious games.” Serious games have proven their value in raising awareness and training. They allow participants to become “actors,” giving them the opportunity to experiment in a risk-free environment. In addition, their playful aspect helps to strengthen motivation and engagement. Each of the games designed as part of SwITCh has its own specific features and aims to raise awareness of different issues related to mobility and/or targets a specific audience.

Analysis of interviews with experts and scientific literature has highlighted the existence of two visions: one driven by technical and technological developments (electric and autonomous cars, connected infrastructure, etc.), and the other emphasizing levers related to the issue of transportation (the 15-minute city, energy efficiency, etc.). Both experts and scientists agree on the need to reduce the use of private cars and make more room for active or semi-active modes of transport. Another common point is the importance given to evaluation and data. This justified the development of the MobiVal4All model.

MobiVal4All is an urban mobility assessment model based on categorized indicators that provides a representation of urban mobility by integrating all its dimensions. This model was designed to be usable by all stakeholders, including non-experts. It was decided to evaluate the indicators on a qualitative scale (from very low to very high), which is easy to understand and comprehend for all types of stakeholders. The model contains a set of 112 variables to calculate 51 indicators belonging to 10 categories. It has been implemented in many French cities and tested with various stakeholders (both experts and non-experts in the mobility system).The model can be used to evaluate prospective scenarios. The method was tested through five participatory workshops and five interviews, resulting in the creation of 21 evaluated scenarios.

To help evaluate the scenarios, two different Agent models were developed as part of the Switch project: Switch_traffic and Switch_modality. The first model aims to simulate urban mobility within a metropolis. It allows for the integration of parametric agendas (modulation of agents' activities according to the scenario) and combines macro and micro dynamics (multi-level) to speed up calculation time. The second model focuses on modality choices; it integrates the evaluation and decision biases specific to this type of decision. The two models were designed to work together, using identical variables and based on compatible structures. They both enable the identification and evaluation of changes in the implicit variables of prospective scenarios.

To reinforce learning, three games have been developed based on these models: a computer game (Sim-switch), a board game (Mobi-switch), and a game for children (Mini-switch). The three games presented are intended for different contexts of use (alone or in a group, independently or with guidance) and require different materials and varying degrees of involvement on the part of the players.

Several obstacles that emerged in the SwITCh project are still present and could be the subject of future research.

Firstly, a difficulty was revealed in the link between the scenarios constructed with the participants (through workshops or interviews) and their evaluation via MobiVal4All. A procedure was developed to automate this process by identifying variables and trends in the narratives. However, some tasks still require human intervention, which requires a certain level of expertise and can be tedious. Rapid advances in AI, particularly conversational AI, are opening up interesting avenues for completely automating this step.

For the MobiVal4All model, one obstacle concerns data. The model has been applied to French cities, which have a large amount of public data available for evaluating the model's various variables. However, this is more complex in smaller urban areas and/or in many other countries where such data is not accessible. This raises the question of missing data in the models and the integration of epistemic uncertainties.

For the simulation model, two obstacles remain: scaling up to the urban area with the computing time that this presupposes; and integrating the different modes into the traffic simulation (current and future modes). Simulation time is a key factor, especially since the proposed models are stochastic and therefore require numerous simulation iterations to converge the results. The question of different modal choices is very important in this forward-looking perspective, in which participants may be led to imagine new modes that would need to be integrated into the models (mode choice and traffic simulation).

Furthermore, as part of the validation of the simulation model developed, a replication was carried out in the proprietary Aimsun solution, modeling the road network of a common case study at two levels of detail. This approach enabled a sensitivity study to be conducted based on indicators commonly used in traffic engineering (speed, travel time, flow). However, the multiscale validation approach still represents a scientific obstacle that will be the subject of future investigations in other research projects.

Transport infrastructures play a large part in defining the city of the future, which should be smart, sustainable and resilient. Their management will need to deal with the emergence of novel technologies (i.e. autonomous cars, Internet of Things) and the increase of novel modalities and practices (increase of multi-modality, electric bicycles, shared cars). These aspects could favour and accelerate the transition to the city of the future with positive social, environmental and economic impacts, in order to address foreseen trends (climate change and new requirements in terms of pollution, security, and global costs).
The SwITCh project aims at supporting decision-making for urban planning by simulating the gradual introduction of disruptive innovations on technology, usage and behaviour of infrastructure. It requires providing a model that is able to assess the impact of these innovations on several key indicators on mobility, user satisfaction and security, economic costs and air pollution. SwITCh integrates a large variety of urban transport modalities (private car, walk, tramway, etc.) and associated infrastructures (pavement, bicycle path, etc.). Achieving such an objective requires building a model that includes current and future infrastructures and modalities, and considering the transition process between current and future situations. SwITCh uses agent-based modelling (ABM) and participative simulation as a unifying framework that allows coupling different models and taking into account both temporal and spatial scales in order to build a holistic model. It will include a city model based on real geographic data (GIS) and a complex realistic model of population behaviour. The model will be designed as a support tool for helping stakeholders (i.e. decision-makers, managers, technicians and citizens) to enrich their reflection and build a shared project to improve transport infrastructures to meet the challenges of future cities.
The SwITCh project will be centred on the design and on the implementation of an ABM that will result in an interactive simulator and a serious game. The interactive simulator will be used by the city planners to explore the potential impact of innovations in various evolutionary contexts. It will thus support the urban planning team in making relevant decisions regarding the evolution of their transport infrastructures, by letting them test and assess different alternatives and situations. The interactive simulator will also allow the researchers to highlight potential futures or unexpected side effects to the urban planners and other stakeholders, based on a participatory simulation approach. The serious game will be used by students and the larger public in order to enrich their understanding of the issues involved in the city of the future and the transport infrastructures. It will be based on the interactive simulator but will be enhanced by specific work on the game design in order to be a real support for learning and raising awareness. The interactive simulator and the serious game will be developed with the GAMA open-source platform and will be used in a real context for two case studies: Bordeaux Metropole and the Urban Community of Dijon.
The SwITCh project will deliver several main results. Firstly, it will generate and formalize knowledge on future transport infrastructures. Secondly, the project will result in a simulation tool that could have significant socio-economic impacts: by helping infrastructure managers and urban planners, as a reflection support, to adapt infrastructures to future needs, by accelerating the transition to a more sustainable city which should have positive environmental (e.g. air pollution, global warming), economical (e.g. maintenance cost, commercial appeal) and social (e.g. traffic, living environment) impacts. The model will be flexible, easily adaptable to any city, and able to integrate a wide variety of prospective and disruptive scenarios.