Railroad Monitoring Drone System Unveiled at Conference

A new system that uses drones to monitor railroads has been presented at the ninth Conference on Railway Innovation.

The goal of this new development is to recognize the good condition of train track elements in images taken by drones. The drones fly over the railroad corridor by means of new image recognition software created by SigmaRail in collaboration with the research group at the UC3M Intelligent Systems Laboratory. This information is then integrated in SigmaQ, a platform that permits accessing the digitalized corridor.

"It is like a kind of Google Maps for railroads that allows the geolocation of trains on the tracks," said Mario Fernández Marín, a co-founder of SigmaRail. "For a train that travels at 300 km an hour, it is very important to know whether a signal, a sign or a curve is 15 m ahead or behind."

The system also detects possible incidences or obstructions on the tracks. To be able to recognize the elements installed in the railroad corridor, it is necessary to manage a large quantity of information. For example, a 20-minute flight for each drone entails 2 GB of data. If each unmanned drone operates for six hours a day, terabytes of data are generated every week that require streamlined processes and cost reductions.

"The definition and refinement of our algorithm of image recognition permits automation of all these processes," said Norberto González Díaz, a co-founder of SigmaRail.

This automation translates into a saving on costs and an increase in safety for users and workers. For example, workers will no longer need to access the railroad corridor to check the condition of the tracks because the software would detect any modification.

The company has flown the technology over a railroad corridor in Spain through a project with Administrador de Infraestructuras Ferroviarias, the Administrator of Railroad Infrastructures. SigmaRail participates in the European Space Agency’s BIC Autonomous Region of Madrid-UC3M Node program, a company incubator project of the ESA managed in Madrid by the madri+d Foundation with the support of the regional government.