10.03.2026

Activity-Based Transport Modelling: Why It Produces Better Urban Plans

When a city wants to understand how a new tram line, a congestion charge, or a park-and-ride scheme will affect travel demand, it runs a transport model. The quality of that model — and therefore the quality of the decision — depends heavily on what kind of model it is.

The dominant paradigm for decades was the four-step model: trip generation, distribution, mode choice, and assignment. It is well understood, widely supported by commercial software, and continues to be used in many cities. But it has fundamental limitations that make it poorly suited to modern transport planning challenges.

The Problem with Traditional Four-Step Models

Four-step models work with aggregate flows — they predict how many trips will be made between zones, not who makes them or why. This creates several blind spots:

• No trip chaining — the model cannot represent the fact that a commuter might drop children at school, stop at a supermarket, and attend the gym — all as part of a single day's travel pattern
• Poor sensitivity to policy — nudges like flexible working hours, MaaS subscriptions, or mobility credits are hard to represent because the model has no concept of individual scheduling decisions
• Limited temporal resolution — daily or peak-hour aggregates miss the dynamics of when and why travel demand shifts
• No within-household interaction — shared cars, childcare responsibilities, and coordinated travel are invisible

What Activity-Based Models Do Differently

Activity-based models (ABMs) start from a fundamentally different premise: travel is derived demand. People do not travel for its own sake — they travel to participate in activities (work, education, shopping, leisure). A well-designed ABM simulates the full daily activity schedule of each individual in a synthetic population, then derives the trips that result from that schedule.

This gives planners several capabilities that four-step models cannot provide:

• Sensitivity to new mobility services — how does shared car access change activity scheduling for households without a private car?
• Equity analysis — which population groups are most affected by a new mobility policy, and does it improve or worsen access for disadvantaged communities?
• Long-horizon forecasting — how might land use changes interact with transport demand in 2035 or 2050?
• Intermodal trip chains — modelling the full door-to-door journey including walk, cycle, transit, and rideshare legs

HarmonyMS and the Activity-Based Modelling Module

HarmonyMS, developed by MobyX as part of the EU H2020 HARMONY project, includes a built-in activity-based modelling module as a core component of its multi-scale planning framework.

What makes the platform distinctive is its software-agnostic architecture: cities that already use commercial tools like PTV VISUM or Aimsun can connect them to HarmonyMS through standardized adaptors, rather than having to replace their existing investments. The platform handles the data flows between models at different spatial scales — strategic, tactical, and operational — while the activity-based module provides the demand side.

HarmonyMS has been applied in:

• Rotterdam — freight and passenger scenario modelling
• Oxfordshire — sustainable mobility planning for the county council
• Turin — multimodal impact assessment
• Athens — urban mobility intervention scenarios

Beyond Planning: System Dynamics for Long-Term Impact

For questions that go beyond traditional transport modelling — such as the system-wide impact of connected and automated vehicles by 2040, or how a mode shift affects public health outcomes over a decade — MobyX also develops IAMT, a system dynamics model designed for long-term societal impact assessment.

IAMT can answer questions like:

• Will increased shared mobility improve or worsen social equity?
• What land use changes might follow from major public transport investment?
• How could connected and automated mobility impact emissions by 2040?

It is used to support EU R&D collaboration, policy white papers, and strategic foresight — representing MobyX's most advanced research capability.

Feeding Models with Real Behavioural Data

Activity-based models are only as good as the data that calibrates them. Smartphone-based travel surveys — collected through platforms like MobyApp — produce the multi-day, individual-level travel and activity diaries that ABMs require. The combination of high-quality data collection and sophisticated modelling is what enables genuinely evidence-based transport planning.

Talk to the MobyX team about how activity-based modelling can support your city's transport planning or SUMP development.