The Norte Litoral Highway runs along the northwest coast of Portugal, from Oporto to Caminha, near the Spanish border, with a branch towards the interior between Viana do Castelo and Ponte da Lima. The highway is 119 km long, of which 72 km are refurbished and 47 km new construction.
As the replication pilot within the Smart Highways domain, this pilot will be executed in close conjunction with the same partners of the pilot domain in order to share the maximum of information and best practices learnt during the initial pilot. However, Norte Litoral adds some additional challenges, since it runs under free-flow infrastructure model, where ITS costs and OPEX are more relevant than in initial pilot (toll infrastructure model in southern coast of Spain).
Pilot will focus on the main big data analysis solutions (clustering, data mining, regression models, machine learning, etc.) and linear programming applications to optimize the highway operation:
- Descriptive modelling of road traffic: Application to increase the understanding of the multivariable nature of highway usage, incidents, and associated phenomena such as black spots and animal run overs.
- Predictive modelling of infrastructure use: Applications to shorten reaction time to incidents and to optimize the infrastructure resources - specially maintenance tasks and any activity under SLA restrictions.
New technological solutions are required to complete the gaps and uncertainties related to the traditional data collection methods in order to obtain reliable, affordable and updated information. Norte Litoral pilot will exploit not only the datasets already available, but as well new data sources given by new ITS devices:
- Distributed acoustic sensing (DAS) based on the existing optic fibre buried along the road shoulder, a solution that allows traffic control centres to receive real-time warnings when a vehicle hits a safety barrier or hits animals on the road.
- Thermographic video analysis system, which is capable of detecting obstacles even in complete darkness as well as through fog or intense rain, complementing DAS systems in identifying pedestrian intrusions, vehicles that have stopped on the road or unauthorised parking.