About

Presentation of the project 

The HARMONIC project “Harmonization of Actions and Reforms on Mobility and Organization in New Interactive Cities,” focuses on evaluating transport and urban development policies. The project's primary goal is to ensure that these policies are better calibrated and harmonized.

A sustainable development strategy must balance environmental, social and economic objectives. However, policies that advance one goal may harm others. This is especially true for mobility policies. For example, low-emission zones (LEZs) reduce air pollution, providing environmental benefits, but they may also exacerbate social inequalities by disproportionately affecting low-income populations who are more likely to own older, non-compliant vehicles. These conflicting impacts highlight the need for adjustment and fine-tuning of policies. Furthermore, different policies may produce adverse effects. For example, promoting electric vehicles or autonomous vehicles could lower the average and marginal costs of travel, influencing household location choices. This might encourage urban sprawl as households look for more affordable and spacious (and thus less energy-efficient) homes farther away from the city centre. In contrast, LEZ policies could induce urban densification for households locating closer to public-transit hubs. Therefore, harmonization between policies is crucial to ensure that they collectively support sustainable development goals.

The goal of HARMONIC is to analyse the short-run (e.g. automobile ownership or mode choice) and long-run (e.g., household and firm locations) impacts of various policy combinations. The period of study is 2025-2050. Our analysis will rely on METROPOLIS2, a dynamic agent-based traffic simulator, and EUrbanSim, a dynamic land use model. We previously developed these unique tools, and will extend and interact them, along with several other tools developed by HARMONIC partners. The focus will be on the cross-effects of different policies and trends. For example, how is the rise of electric vehicles compatible with the introduction of new public transit lines? These policies have been evaluated individually. A radial public-transit line can have centripetal effects, while a reduction in kilometre costs (as is the case with electric vehicles compared to thermal vehicles) can lead to urban sprawl. Similarly, the zero net land artificialization (ZAN) legislation can conflict with other policies, such as the development of transport infrastructure: the construction of new tramway, metro, or bus lines can sometimes require land artificialization, especially on the outskirts of cities. Densification can also lead to tensions regarding quality of life, green space management, and public services. To sum up: combining economic development with decarbonization of the society remains a huge challenge.

To meet our policy harmonization challenge, we will build an operational dashboard for evaluating the combined effects of a wide range of transport and urban policies. The dashboard will facilitate discussions with stakeholders to determine the best ways to harmonize policies, taking into account the short, medium and long run time scales.

Project outline

The project is structured into 32 interconnected subtasks (detailed in Table 1), grouped in 8 tasks (including dissemination and coordination), as briefly described below.

Data Collection (Task 1): Collection of a wide range of data on population, firms and land use data, plus an original survey on urban freight and weekly mobility.

Synthetic Population (Task 2): Building synthetic populations of individuals, firms, buildings, private vehicles, and trucks evolving over time, using Task 1 data. All synthetic populations are used both in our simulators METROPOLIS2 and EUrbanSim, and for illustrating our theoretical developments.

Policies (Task 3): Design (to feed simulators), evaluation (using our simulators outputs), and harmonization of various policies such as electric cars, automated vehicles, SERM, low emission zones, and zero net soil artificialization, to ensure the optimality of their interactions.

Model Development and Calibration (Tasks 4 & 5): Further developments and adaptation of our transport and land use models, including evolving preferences and automatic calibration. METROPOLIS2 and EUrbanSim are integrated, and jointly used to evaluate policies.

Dashboard (Task 6): dashboard engine and interface to manage inputs/outputs of policies.

Dissemination (Task 7): Dissemination of the results and insights gained from the project.

Coordination (Task 8): Ensuring effective coordination of our highly interconnected tasks. The main study areas of HARMONIC will be Île-de-France and in Strasbourg Métropole.

Scientific strategy

The framework of our work is straightforward: data collection, policy definition, simulation using an integrated tool for transport and land use, representation of policies affecting short-term choices (weekly mobility) and long-term choices (urban development). Our main ultimate goal is to identify ways to harmonize policies to avoid redundancies and contradictions and to improve synergies. We add various crucial complexities: (1) considering decisions within families; (2) accounting for changes in preferences; (3) considering mobility on different days of the week; (4) integrating freight transport.

The main innovative scientific contribution of HARMONIC is the integration of different tools in a unified framework combining data production and collection, theoretical developments (micro and public economics), econometrics (structural models), data science (for the automatic calibration), and engineering (on the supply side). The synthetic populations will offer a bridge ensuring consistency between the different simulators, and allowing to compute the behavioural and redistributive effects of the policies evaluated and harmonized, both in the short run and in the long run. Crucial guidelines will be provided by the stakeholders and other practitioners, and by the field knowledge of several participants in the consortium. The consortium has demonstrated its expertise in each of the component tools, models and subject areas in HARMONIC. Each component will be fine-tuned and enriched in order to better integrate it with the other components. Consortium participants who are not specialized in a given component will bring fresh eyes and insights to aid the specialists.

Challenges

At first glance, HARMONIC may look too ambitious. However, the main challenge is not to perform individually each of the developments considered, but rather to integrate them. Our first challenge is thus to ensure that the work of each team/discipline aligns with the needs and expectations of the others. Our numerous past collaborations, and the careful management of the project (see 2.1.2) will allow us to meet this challenge. Drafting this proposal was our first exercise in alignment. Each of the 4 items listed in 2.1.3 is innovative and, as with any research, contains some uncertainty. Nonetheless, with specialists in each of these areas on our team, the major remaining challenge will be to integrate them into a unified framework and demonstrate the benefits of studying them. The consortium as a whole offers a spectrum of solutions large enough to address each specific source of uncertainty or unexpected obstacle. Some specific technical challenges may be mentioned. We rely on two main models: the land use model describes the dynamics of the long run evolution of the urban system (slow adjustment process); the traffic model computes short run stationary solutions (fast adjustment process). The challenge is thus to reconcile these two fundamentally different time scales. We will meet this challenge by consistently integrating the two models, using synthetic populations carefully built for this purpose. When evaluating the long-run effects (typically up to 2050) of transport and urban policies, account should be taken of changes over time in both preferences (e.g., attitudes towards non-motorized modes) and technology (e.g., teleworking, electric cars and autonomous cars). To meet this challenge, we will use panel data techniques to measure the trends of such evolutions and, where needed, exploit experimental economics data already collected by HARMONIC partners. Finally, the quality of the results depends on the quality of the data. In Task 1, we rely mainly on open data (national and local travel surveys, OpenStreetMap networks, Census, TomTom free access, and land occupation data) and on data collected internally (for freight and weekly activity).

Rationale for this consortium

We established an interdisciplinary and highly complementary consortium including economists (BETA and ThEMA), computer scientists (LIVE and ThEMA), geographers and physicists (LIVE), land planners (IPR and MATRiS), engineers (LVMT and SPOT), industrial engineering (BETA).

Three external participants (N. Geroliminis, P. Stokkink, R. Lindsey) will strengthen our international scientific excellence. Most participants have already worked together and/or published together, which will make it easier to break down the interdisciplinary barriers, usually difficult to overcome.

Scientific and technical description of the project

Figure 1 describes the four modules of HARMONIC: (1) Synthetic populations and activities for freight, private, and public transportation, including decision trees for ownership, trip chaining, and family-level itineraries; (2) Simulators for transport, land use, and pollution: develop, calibrate, and run our tools; (3) Policy scenario construction, evaluation and harmonization; and (4) Dashboard for presenting results. The corresponding (sub)tasks are described below and detailed in Table 1.

Expected outcomes of the project

Scientific expected outcome

Publications of articles in transport economics, urban economics, geography, regional sciences, energy, environment, social choice and policy analysis. We aim to publish at least eight articles in top-tier journals, with two or three in prestigious publications (Science, Nature, Nature Communications, Nature Climate Change or PNAS). Other papers will be aimed at top field journals. We will write 2 position papers (see 3.3.3). Articles will be submitted to The Conversation, or as IPR’s Note rapide. Our results will be accessible to policymakers through our dashboard (see details in section 2.2.6). We will also produce a few videos, accessible from our website and YouTube aimed at policy makers and at an educated general public. In section 3.3.3, we describe the valorization of our activities.

Valorisation and Transferability to other French cities

The bi-monthly online transport and land use seminar, originally established as part of the ANR AFFINITE project, will be renamed HARMONIC seminar. It is open to students and faculty in France, Europe, and North America. This seminar will feature ongoing research by project participants, as well as by other researchers working on similar topics. The presentations will be recorded, and sufficient time will be allocated to discussion (the seminar lasts 1.5 hours, including questions). Starting in year two, we will offer an online course on METROPOLIS. Starting in year three, it will be completed supplemented by an online course on urban modelling. The two courses with be consolidated in the last year (with possibly the organization of a summer course, to meet potential demand). The idea is to: (a) teach young scholars and practitioners how to use our tools, and (b) identify future partners who can use our platform for other cities. A trial online course (10 3-hours sessions), attended by 15 international students was offered by L. Javaudin in Spring 2024, for METROPOLIS2.

The automatic calibration of METROPOLIS2 and EUrbanSim, along with their integrated version, will enable within a few months to perform simulations for any city or region in France. The tool primarily relies on synthetic populations generated from open data available for all French cities. With minimal additional work, policy evaluations could be extended beyond Paris and Strasbourg to other cities. Throughout the project, we will identifying local partners and encourage collaborative efforts in these directions. Our goal is to create a platform accessible to other interested parties. This process is currently more standardized for the short-term traffic model than for the urban long-term component.