SiR, smart processing for smart railways Transport is a fundamental pillar of our economy and society, enabling economic growth and jobs creation. However, the railway industry must face critical challenges to guarantee passengers and freight safety as passenger's mobility increases (200-300% by 2050) and freight activity grows. Among these challenges, the efficient and successful management and operation of infrastructures is key for the improvement of the services offered by railway infrastructures administrators and operators. Maintenance, both active and predictive, is one of these major challenges. The railway industry is addressing it by putting in place a process of digitalisation. The advent of new sensors and technologies, such as 3D scanners and LiDAR, is helping the industry to cope with the first step in this process, collecting digital models from reality. However, the vast amount of data produced makes manual data analysis not practical. SigmaRail proposes to develop SiR, a deep learning and neural network-based system that will automatically process collected data and will provide useful insights of the railway infrastructures and assets status. Short term benefits include the improvement of the active maintenance services, as well as sets the base for the development of predictive maintenance analysis. Passengers and freight transportation will be notably safer and more efficient. This project intents to develop an innovative system, to apply forefront technologies such as neural networks to specific railway industry use cases.