What if Europe’s highways can be smarter, safer and more efficient? Two pilot projects under the formidable EU-funded Transforming Transport project – which boasts 47 organisations from 9 countries and a budget of EUR 18.7 million – are currently ensuring that this vision becomes reality. The Smart Highways pilots, as they are known, are aimed at understanding traffic flows and mobility patterns in order to optimise highway operations and take necessary action based on predicting traffic behaviour. The pilots are also aiming to reduce the number of accidents due to traffic and animal intrusions.
Overall, the pilots will measure the technical and economic impacts of Big Data applied to the highway markets operated by Cintra, one of the world’s leading companies in financing
and managing transportation infrastructure based in Spain. There are two pilots: the first applies to the Ausol Highway (Spain) and a second one applies to Norte-Litoral Highway (Portugal) that will replicate the Big Data solutions deployed in the Ausol case.
Big Data technology improves Spanish Highways
The initial pilot is the 105 kilometre Ausol highway: this highly-congested semi-urban corridor connects the cities of Málaga, Estepona and Guadiaro in the South of Spain. The AP-7 highway is part of the European route E15, Costa del Sol and the alternative road is the semi-urban N-340/A-7 road.
One of the premises and the most valuable contribution to this project is to be able to get the data in real time from the data sources which the pilots are working with. To do this, the pilots developed an online “dashboard” for displaying information to Traffic Control Centre operators in a user-friendly manner. The dashboard exploits the cloud and can be accessed through laptops and mobile phones, displaying traffic graphs, CCTV camera feed and traffic alerts. Importantly, the dashboard’s graphs show historical, real-time and forecasted traffic data for the next two hours. Some of the crucial data that the dashboard provides includes real time maintenance on highways and real time queue length at toll stations.
The system is capable of calculating transit times for each toll lane in real time, depending on the type of payment. An alarm warns of abnormal queues at the toll station, helping to optimise resources and reducing waiting times to improve user experience.
Big Data technology upgrades Norte Litoral’s Highway
The replication pilot runs on the 119-kilometre infrastructure of Norte-Litoral highway. This corridor connects the cities of Oporto, Caminha and Ponte da Lima in the North of Portugal. The A-28 highway features an electronic toll system that is completely free from obstacles for users.
As in the initial pilot (Ausol), the pilot developed a dashboard that has been adapted to both the data available and the needs of the highway.
This pilot differs from the previous one in that there are no toll plazas and no data from the tolls. It relies instead on gathering information from traffic data, traffic events, twitter, weather data, travel times, socio economic, and messages from computerized road signs (PMV).
In addition to all data sources, technology currently used in the railway industry is being deployed at Norte Litoral. Known as Distributed Acoustic Sensing (DAS), the technology uses the existing optical fiber buried along the highway to detect patterns on the road vibrations to deduce odd behaviors that might affect traffic flow. In the future, this will hopefully represent a new source of data capable of detecting in real time incidents that occur on the road, reducing response time to traffic-related incidents.
Descriptive Data Analysis
An important feature of the pilots is that they offer both a predictive traffic model and a road accident travel model. In the predictive traffic model, the pilots forecast not only based on the 120-minute approach but also a 15-minute approach used to double-check the 120-minute one and adjust it if necessary.
In parallel, the road accident traffic model, currently under development, maps one-to-one relationships between accidents and different variables that could cause them. In certain occasions, it has undertaken deeper analysis, cross-referencing accidents with two variables at the same time. This helps reveal whether a variable is directly related to an accident and helps decide if the variable is to be included or discarded in the model.
The dashboards and technologies developed under these two pilots are without a doubt important contributors towards advancing the transformative effects that Big Data will have on the industry.