Machine learning and artificial intelligence have showed in the last decade their strength to solve complex problems difficult to face with conventional algorithms or only with simulation techniques1,2. The fast increase of the digitalization and the large amount of available data that can be obtained from several sources, makes artificial intelligence the most suitable tool to work with complex problems, such as the traffic management.
In FRONTIER, AI will be used to develop models able to predict (detect prior its occurrence) incidents in the traffic network and determine the consequences in the related areas. This will allow at the same time to determine the optimal path to avoid traffic jams by using simulation techniques, which will use as input data the outputs obtained from the AI models. The platform developed in the FRONTIER project will be also used to perform an optimized management and communication of the full traffic network taking into account all the different agents involved, like pedestrians, bikers, metro, buses, drivers and maritime transport among others.
To set up the AI models and test their efficiency three pilot areas located at Oxfordshire (UK), Athens (Greece) and Antwerp (Belgium) will be used. The challenge in the UK, thanks to the current work of the Oxfordshire County Council (OCC), will be focused on the study of the Connected and Automated Vehicle (CAV) to evaluate its influence in the traffic system and determine the best practices for its integration. In Athens, the FRONTIER project will test the intelligent platform in the 70-Km-long Attiki Odos motorway in collaboration with the Athens Urban Transport Organization (OASA) and Attiko Metro (AMETRO). It is expected to obtain an efficient interconnection between all the traffic agents to obtain an optimized management, allowing a resilient and secure traffic system. Finally, the port city of Antwerp will be used to study and develop models considering interactions between maritime and land transportation to relief congestion in specific areas.
Figure. Attiki Odos in Athens (Left), CAVs in central Oxford (Center) and boat in Antewerp Port (Right). Sources: N. Danilidis, StreetDrone e-NV200 and Streed Drone Twizy and UA picture database
Together with the AI models developed from large sets of data gathered from different sources, including the pilots, the Eurecat Technology Center will also contribute with the implementation of a semantic model built from a specific ontology based on smart cities and the automotive sector will be used to provide more context information and to support data analytics and visualization.
In conclusion, it is expected that in the context of the FRONTIER project the artificial intelligence will help to setup the basis of the future smart mobility, not only from the point of view of the efficiency, resilience and safety of the traffic network, but also considering the business models of all the related stakeholders towards a sustainable economy, building a green mobility model for the coming generations.
 Zaib Ullah, Fadi Al-Turjman, Leonardo Mostarda, Roberto Gagliardi, “Applications of Artificial Intelligence and Machine learning in smart cities”, Computer Communications, Vol 154, 2020, pp. 313-323.
 Akbari, M. and Do, T.N.A. (2021), "A systematic review of machine learning in logistics and supply chain management: current trends and future directions", Benchmarking: An International Journal, Vol. 28 No. 10, pp. 2977-3005.