MCRIT works with the Spanish Ministry of Transport on the macro-simulation of passenger and freight transport flows across Spain for different transport modes. It also works with Renfe and Adif on rail transport. Transport models estimate future demand and service levels across transport networks. To do this, they consider changes in socioeconomic conditions, infrastructure supply and management, public transport services, land use, and urban development.

MCRIT develops these macro-simulation models for the Ministry of Transport using PTV VISUM software. This software applies transport economics theories and algorithms. Their foundations date back to the first models developed in Detroit, USA.

From data scarcity to data abundance

In the past, the main challenge in developing transport macro-simulation models was the lack of information. Surveys and traffic counts were expensive, which made data collection difficult. Today, the challenge is different. We now need to manage massive real-time databases and develop large-scale microsimulation models.

Medium- and long-term macro-simulation models are known as “four-step models”. These include trip generation, distribution, modal split, and assignment. Over time, these models have evolved to include land-use effects and social and economic activities. Data from mobile phones, sensors, and other sources now allow teams to build models based on behavioural patterns. In some cases, these patterns can be individualised. Artificial Intelligence can also support the statistical calibration of complex algorithms.

Planning, forecasting and expert knowledge

These new generations of models can greatly improve the short-term management of public services, especially transport. However, infrastructure planning still requires medium- and long-term scenarios. For these scenarios, current information alone is not enough. Expert knowledge must complement it.

Traditional four-step models therefore remain the most widely used. In microsimulation models or real-time management systems, it may be useful to let the data speak for themselves. In macroscopic models, however, engineers still need strong planning expertise and must retain the final say over the model. In the short term, the main goal is accurate forecasting. In the medium and long term, the main goal is understanding.