ERC Consolidator Grant 2025 · 60 months

Transport resilience for an uncertain world.

TRANSFORM builds a smart estimate-then-optimize framework that connects real-time decisions to long-term investments in multimodal urban mobility — so cities can adapt to disruption instead of merely react to it.

Methodology →Contact us

Hosted at TU Delft · Dept. of Transport & Planning

Uncertainty

Dynamic, multi-source quantification of disruptions — from accidents and infrastructure failure to extreme weather — embedded directly inside the optimization loop.

Behaviour

User-centric multimodal trip menus, with choice parameters that adapt as new mobility solutions enter the network — no surveys, no biased re-estimation.

Resilience

Iterative optimization that bridges operational, tactical, and strategic timescales — so today's disruptions inform tomorrow's investments.

What is TRANSFORM?

Multimodal passenger networks are under constant pressure from new technologies, shifting business models, and the simple fact that more people are travelling in denser urban space. Yet today's transport systems are designed in silos: real-time disruption insight rarely informs long-term investment, and strategic plans rarely flex with daily reality.

TRANSFORM is a five-year research programme that develops the first scientifically grounded framework to bridge those silos. It coordinates three players of the mobility system — operators, infrastructure authorities, and travellers — through three integrated pillars: dynamic uncertainty modelling, behaviourally informed demand management, and iterative cross-timescale optimization.

The work is hosted at TU Delft and led by Associate Professor Shadi Sharif Azadeh, with a team of three PhD researchers and two postdoctoral fellows.

The Framework

One horizon. Three coordinated objectives.

Traditional transport planning treats minutes and decades as separate problems. TRANSFORM unites them on a single decision axis.

Fig. 1 · A unified decision horizon

One framework — from seconds to decades.

From real-time disruption response to multi-decade infrastructure planning — all three TRANSFORM objectives live on the same decision horizon.

OperationalTacticalStrategic

Research Objectives

What we will deliver.

Three interlocking objectives — each closing a gap in current transport science.

Short term · seconds → hours

Multimodal supply management under uncertainty

Real-time probabilistic forecasting of resource availability — fleets, vehicles, chargers — that integrates directly into a centralized optimization for service scheduling and dispatching.

Bayesian forecastingService schedulingFleet management

Short term · individual scale

Demand management & behavioural adaptation

A user-centric multimodal trip recommender built on assortment optimization. Choice parameters are re-calibrated whenever new mobility services appear — using LLM-enriched data and simulation-based population synthesis, without rerunning surveys.

Discrete choiceAssortment optimizationLLM-enriched data

Long term · months → decades

Network adaptation & expansion under uncertainty

Interpretable probabilistic predictions — built on Bayesian networks — capture causal effects between established and new mobility services, feeding a strategic adaptation model that respects daily operational reality.

Bayesian networksAdaptive optimizationNet-zero investment

Work Packages

How the research is organised.

Four work packages — three vertical objectives plus a cross-cutting uncertainty workbench.

WP0

Uncertainty quantification & impact assessment

Cross-cutting work that identifies uncertainty sources and powers TRANSSim, a new multi-agent simulator extending MaaSSim with network adaptation and centralized trip recommendation across temporal scales.

PD1 · PD2 · PI

WP1

Multimodal supply management under uncertainty

Two-stage Bayesian-inspired forecasting of resource availability, integrated with a real-time optimization framework for centralized service scheduling.

PhD1 · PD1 · PI

WP2

Multimodal demand management under uncertainty

Choice-parameter adaptation against environmental change plus a trip recommender that steers demand toward fleet availability via tailored multimodal menus.

PhD2 · PD2 · PI

WP3

Long-term multimodal network expansion

A robust estimate-then-optimize framework that connects short-term operational evidence to strategic adaptation under deep uncertainty.

PhD3 · PD2 · PI

Simulation Platform

TRANSSim — a multi-agent testbed for the whole programme.

TRANSSim extends the open-source MaaSSim platform with two new layers: a network adaptation component and a centralized multimodal trip recommender. A discrete event scheduler manages agents acting on very different clocks — hourly fleet rebalancing alongside annual infrastructure decisions — keeping the whole system temporally coherent.

Use cases will be calibrated on multi-source mobility data collected from the Netherlands, Germany and Norway, with open-access data sets used as a back-up when proprietary feeds aren't available.

TRANSSim · live preview

Project Impact

Why it matters.

Three groundbreaking contributions across transportation science, behavioural modelling, and long-horizon network design.

A paradigm shift in transport management.

The first scientifically grounded framework to connect short-term operations and long-term planning, replacing reactive decision-making with proactive, dynamic uncertainty modelling.

Behavioural models that adapt without resurveying.

Choice parameters that update themselves when new mobility solutions enter the network — leveraging LLM-enriched data and simulation-based population synthesis to save both time and cost.

Foresight for systems that span six decades of literature.

Vickerman (2024) and Foster (1964) asked the same question — how to expand transport in a balanced, equitable way. TRANSFORM answers it with interpretable probabilistic predictions that survive incomplete data on new services.

Use Cases

Where TRANSFORM is being tested.

Multi-source mobility data has been secured across three European cities.

Netherlands

Rotterdam

Coordinated multimodal recommender systems with the public-transport operator, integrating fixed-line transit with shared micromobility.

Germany

Munich

Disruption-resilient multimodal scheduling, with a focus on the propagation of delay across rail, bus, and shared modes.

Norway

Oslo

Long-term network adaptation under weather-driven uncertainty, including electrified fleets and mobility hubs replacing legacy refuelling sites.

Principal Investigator

Leading the programme.

Principal Investigator

Shadi Sharif Azadeh

Associate Professor · TU Delft · Transport & Planning

Shadi co-founded and co-directs the Sustainable Urban Multimodal Mobility (SUM) Lab at TU Delft. Her research sits at the intersection of operations research, transportation engineering, and behavioural modelling — with a track record of bringing choice-based optimization into transport planning.

She holds a PhD in operations research from Polytechnique Montréal, has been a visiting scholar at MIT and Georgia Tech, and was a Marie Curie Postdoctoral Fellow at EPFL. Recent recognitions include the INFORMS Early Career Award and a Meritorious Service Award from Transportation Science.

2023 — now

Associate Professor, TU Delft

2021 — now

Co-director, SUM Lab (TU Delft)

2016 — 2021

Assistant Professor, Erasmus School of Economics

2014 — 2016

Marie Curie Postdoctoral Fellow, EPFL

2013

PhD, Polytechnique Montréal

Team — three PhD researchers + two postdoctoral fellows

PhD researcher

PhD 1

Multimodal supply management & real-time forecasting.

WP1

PhD researcher

PhD 2

Demand management & behavioural adaptation.

WP2

PhD researcher

PhD 3

Long-term network adaptation under uncertainty.

WP3

Postdoctoral fellow

Postdoc 1

Uncertainty quantification & TRANSSim.

WP0 · WP1

Postdoctoral fellow

Postdoc 2

Cross-timescale integration & methodology.

WP0 · WP2 · WP3

News & Updates

Latest from the programme.

Project milestones, paper releases, and team announcements will appear here as the work progresses.

2026 · AprMilestone

TRANSFORM kick-off planned for autumn 2026.

The five-year programme will begin with WP0 and recruitment of the first postdoctoral fellow alongside the three PhD researchers.

2026 · MarOpen positions

Recruitment for three PhD positions opening soon.

Across WP1–WP3 the programme is seeking researchers in operations research, transport modelling, and behavioural econometrics.

2026 · FebPublication

Foundational paper on choice-driven service network design.

A key precursor to WP2's trip-recommender framework — building on a decade of choice-based optimization work.

Selected Publications

Research the programme builds on.

A subset of foundational work by the PI that informs TRANSFORM's methodology.

Electrification of a bus system with fast charging stations: impact of battery degradation on design decisions.

Sharif Azadeh, S., Vester, J. & Maknoon, Y. · Transportation Research Part C

2022

Choice-driven dial-a-ride problem for demand responsive mobility service.

Sharif Azadeh, S., Atasoy, B., Ben-Akiva, M., Bierlaire, M. & Maknoon, Y. · Transportation Research Part B

2022

Choice-driven service network design for an integrated fixed-line and demand-responsive mobility system.

Sharif Azadeh, S., Van der Zee, J. & Wagenvoort, M. · Transportation Research Part A

2022

Integrating advanced discrete choice models in mixed integer linear optimization.

Pacheco Paneque, M., Bierlaire, M., Gendron, B. & Sharif Azadeh, S. · Transportation Research Part B

2021

Integrated timetabling and vehicle scheduling of an intermodal urban transit network: a distributionally robust optimization approach.

Xia, D., Ma, J. & Sharif Azadeh, S. · Transportation Research Part C

2024

Exact and heuristic algorithms for cardinality-constrained assortment optimization under the cross-nested logit model.

Zhang, L., Sharif Azadeh, S. & Jiang, H. · European Journal of Operational Research

2024

Optimization of the location and capacity of shared multimodal mobility hubs to maximise travel utility in urban areas.

Xanthopoulos, S., van der Tuin, M., Sharif Azadeh, S., et al. · Transportation Research Part A

2024

Interested in collaborating with TRANSFORM?

The programme welcomes industry partners, public-transport operators, and academic collaborators working on resilient multimodal mobility. Get in touch with the PI.

Email the PI →Visit TU Delft