In recent years, Artificial Intelligence (AI) has evolved from a futuristic concept to a useful and powerful tool in multiple sectors. In the field of transport and mobility, its application is not only limited to the operational management of traffic in real time (which is one of the most widespread applications of AI in the sector) but is revolutionising all phases of a traffic project, from the starting point, as innovative forms of data collection and processing begin to be applied, through the phases of planning, management, execution and even monitoring of the project.
At Vectio, we have been integrating these technologies in our studies for years, using for example artificial vision cameras to obtain accurate information on real traffic behaviour. Thanks to machine learning, traffic models can incorporate not only counts and gauges, but also behavioural patterns, non-obvious relationships between variables, or responses to structural changes in the urban environment. This is especially useful when assessing how a new development (such as a new logistics space, residential area or retail space) will affect mobility in the surrounding area. In the traffic and mobility sector, AI offers different efficient applications for traffic studies, and today we would like to review some of the tools that we consider most useful, that are already established in traffic and transport planning, and that we use in the drafting of our studies.
- Data collection and processing: Through machine vision technologies, such as cameras and vehicle detection algorithms, rich and reliable data on volumes, trajectories and modes of travel can be obtained. Although this technology has been applied for years, the great advances in AI tools have made it possible to go a step further, refining data collection and allowing for greater richness in the data obtained, identifying vehicle patterns, movements, anomalous movements, or vehicle models, among others.
- Predictive modelling: AI algorithms make it possible to project future mobility scenarios taking into account variables such as population growth, changes in land use, evolution of demand or new infrastructures. This helps to anticipate conflicts, evaluate solutions and justify investments on a technical basis. In transport engineering, specifically in the estimation of generated trips, four-stage models (generation, distribution, modal split and assignment) have traditionally been applied. The various AI tools available today have made these models evolve, streamlining the processes and integrating them in a backbone manner in all phases of a four-stage model.
- Sensitivity and scenario análisis. In contrast to traditional approaches, which were based on static assumptions, AI allows multiple scenarios to be explored in an agile and automated way. Different road configurations, mitigation measures or mobility strategies can be compared from an integrated view.
- Route and access optimisation. When planning a new development, it is essential to justify safe vehicular access and ensure that the development will not have a significant impact on the surrounding road network. Thanks to AI tools that spatially analyse the environment, efficient access routes can be obtained, which are adapted to the activity to be developed and which allow optimisation of travel times and entrances, without causing additional vehicular delays.
- Monitoring of projects in real time. The use of AI tools, together with the use of Big Data technologies, makes it possible to monitor projects in real time, so that during the execution of the project itself, traffic results can be observed and, in this way, appropriate decisions can be taken and interventions can be made as quickly as possible. For example, in road reconfiguration projects, it is possible to observe the traffic flowing through the environment using Big Data applications, and incorporate an AI tool that analyses and displays the main results, such as possible traffic jams or other indicators of traffic conditions.
In addition to these tools, other more generalist AI applications, such as those using generative language models (Large Language Model) also have a place in the planning phases, as long as they use models trained to be technically reliable, using high quality data or incorporating continuous validation and evaluation processes. However, at present, even if different control mechanisms are implemented, these tools are limited by human supervision, which makes them still highly dependent and diminishes their effectiveness to some extent.
As a main conclusion, the incorporation of AI in transport planning not only improves the accuracy of analyses, but also provides new opportunities, which are particularly valuable in an evolving urban context, where planning decisions must take into account factors such as sustainability, energy efficiency, equity in access to mobility and resilience to uncertain scenarios.
In short, Artificial Intelligence does not replace technical criteria or the experience accumulated over years of work, but rather enhances them. At Vectio, we see AI as a complementary tool that allows us to analyse the environment more efficiently in order to obtain more accurate and appropriate results. As cities and regions are committed to smarter, more sustainable and people-centred planning, having technologies that convert the vast amount of data available today into knowledge becomes essential and, on that path, AI offers a multitude of possibilities to obtain, from traffic and transport planning, more sustainable and appropriate solutions for the world of tomorrow.
Thank you for reading.
Carlos González | Civil Engineer
