Development of a prototype of an autonomous obstacle detection system
The SMART prototype solution for obstacle detection will provide prototype hardware and software algorithms for obstacle detection. The system will combine two night vision technologies, thermal camera and image intensifier, with multi-stereo vision system and laser scanner to create fusion system for short (<20 m) and long range (<1000 m) obstacle detection during day and night operation. By this planned fusion of sensors, the system will be capable, beside reliable detection of obstacles, to provide short range (<200 m) wagon recognition for shunting operations.
Development of a real-time marshalling yard management system
The SMART real-time marshaling yard management system will provide optimization of available resources and planning of operations in order to decrease overall transport time and costs associated with cargo handling.
A web-based information system will be developed for visually representation of the marshalling yard configuration, to provide manual or automated input of inbound and outbound train parameters, as well as planning of marshaling process using the machine learning based optimization algorithm.