We are aware that safety at level crossings is a critical challenge for both rail and road traffic. iCROW is designed to identify abnormal situations and behaviours that could compromise safety using advanced computer vision technologies. The system detects obstacles regardless of their metal base, ground clearance or other existing detection systems, ensuring reliable and precise coverage.
Through image processing and object classification, iCROW can distinguish and record a wide variety of elements, including vehicles, pedestrians, animals and fallen objects. The detection area is defined via a software application and can be adjusted for each level crossing, while the system also monitors any physical changes in the environment.
A processing unit continuously analyses images captured by high-resolution cameras and a communication module issues real-time alerts for any detected obstacle, enabling immediate action to prevent accidents or dangerous situations.
iCROW incorporates advanced technologies that increase its efficiency and reliability:
-
Convolutional Neural Networks (CNNs): to analyse images captured by the AI cameras.
-
Real-Time Remote Monitoring (RTRM): remote monitoring from a control centre to act immediately in case of danger.
-
Machine Learning: algorithms that improve system accuracy by learning from historical and future data.
-
IoT (Internet of Things): a network of interconnected devices that enables real-time monitoring and analysis of level crossings.
The system has already been validated in Spain by ADIF (Administrator of Railway Infrastructures), ensuring its reliability for critical infrastructure and public safety applications.
More than surveillance, iCROW provides active prevention. Its immediate detection capabilities and integration with railway control systems allow incidents to be anticipated, risks reduced and safer, more efficient mobility ensured at railway level crossings and intersections.
