Periodic Reporting for period 2 - SMART (Smart Automation of Rail Transport)
Okres sprawozdawczy: 2018-04-01 do 2019-09-30
The main goal of the project SMART- SMart Automation of Rail Transport is to increase the effectiveness and capacity of rail freight through the contribution to automation of railway cargo haul at European railways by developing of:
• a prototype of an autonomous obstacle detection system, and
• a real-time marshalling yard management system.
The SMART solution for obstacle detection (OD) provides prototype hardware and software algorithms for obstacle detection on the rail tracks ahead of the locomotive. The system combines different vision technologies: thermal camera, night vision sensor (camera augmented with image intensifier), multi stereo-vision system, and laser scanner in order to create a sensor fusion system for mid (up to 200 m) and long range (up to 1000 m) autonomous obstacle detection during day and night operation, as well as during operation in poor visibility condition. In this way, SMART will make an important contribution to the vision of a fully automated rail freight system.
The SMART real-time marshalling yard (RTMY) management system provides optimisation of available resources and planning of marshalling operations in order to decrease overall transport time and costs associated with cargo handling. The yard management system provides real time data about resources available over open and TAF/TSI standard data formats for connection to external network systems and shared usage of MYs between different service providers
• a prototype of an autonomous obstacle detection system, and
• a real-time marshalling yard management system.
The SMART solution for obstacle detection (OD) provides prototype hardware and software algorithms for obstacle detection on the rail tracks ahead of the locomotive. The system combines different vision technologies: thermal camera, night vision sensor (camera augmented with image intensifier), multi stereo-vision system, and laser scanner in order to create a sensor fusion system for mid (up to 200 m) and long range (up to 1000 m) autonomous obstacle detection during day and night operation, as well as during operation in poor visibility condition. In this way, SMART will make an important contribution to the vision of a fully automated rail freight system.
The SMART real-time marshalling yard (RTMY) management system provides optimisation of available resources and planning of marshalling operations in order to decrease overall transport time and costs associated with cargo handling. The yard management system provides real time data about resources available over open and TAF/TSI standard data formats for connection to external network systems and shared usage of MYs between different service providers
The main results achieved within OD workstream in the 1st reporting period involve definition of requirements and specification of the multi-sensory system. Starting from the defined requirements and specification, the sensors of OD system were selected/developed. All sensors were integrated into the sensor housing to enable easy mount and dismount onto/from different test vehicles. Conformance testing of all OD sub-systems was done, and the several field tests were conducted on railtrack sites in Aachen, Germany, and in Nis, Serbia. The HW infrastructure was fully completed and ready for further fireld tests in the next project period to support further development of OD software for autonomous recognition of obstacles on the railtracks ahead of the locomotive and for the calculation of distances to those obstacles.
The main results achieved within OD work stream in the 2nd (final) reporting period involve the final system integration of the on-board OD system as well as the finalization and implementation of the software for OD and distance estimation. The functional testing of the fully integrated system was performed in dynamic field tests and the evaluation testing of the integrated OD system was performed in operational environment. The dynamic field tests for the purpose of functional testing as well as the evaluation tests were conducted on the railtrack sites in Serbia, achieving so TRL4-5. The permissions for the tests in operational environment were issued by Serbian Railway upon submission of needed documentations on the developed OD system and its integration onto operational locomotive.
The main results achieved within MY workstream in the 1st reporting period involve analysis and specification of requirements for MY management. Modelling processes in MYs were done and essential data and processes for providing optimisation of available resources and planning of marshalling operations in order to decrease overall transport time and costs with cargo handling were selected and described in details. Standard data formats were analyzed and were taken into account in the initial phase of developing real-time yard management (RTYM) system. Concept solution was done as well as database structure of the RTYM system. Software development for real time decision support system was started based on achieved results in WP4 and WP5. Initial version of Visualization and Planning module with Data Input and Data Output submodules was done during the reporting period.
The main results achieved within MY work stream in the 2nd (final) reporting period involve software development of advisory decision support RTYM system and development of algorithms for processes in local marshalling yard and their optimization. RTYM mainly targets deviations in decision-making processes in marshalling yard with respect to time and infrastructure. This system offers dispatchers and other decision-makers alternative solutions in case of deviations. The main objective is to provide optimization tool to decrease overall transport time and costs with cargo handling. Providing prompt solutions for real-time deviations in decision-making processes with presentation of potential consequences, SMART RTYM increases efficiency on local marshalling yards and decrease time for making decisions. SMART RTYM system was tested on real-time data provided from two local marshalling yards (from Serbia and Bulgaria).
The SMART project results were disseminated at relevant Conferences and other dissemination events. The SMART project results were presented at several relevant external Conferences in the Railway Engineering, Computer Vision and Robotics fields, as well as at the Conferences organised by SMART partners: RAILCON2016, RAILCON2018 and the first International Railway Symposium Aachen IRSA2017. SMART project mid-term results were presented at the Shift Freight to Rail-Midterm Event and the SMART project final event was at the Innovation in Freight Conference, Munich, 18 June 2019.
The SMART obstacle detection demonstrator was displayed at the Shift2Rail stand at the InnoTrans2018, 18th-21st September 2018.
SMART partners will continue keeping on with dissemination of project results beyond the project lifetime to assure better sustainability and usability of the project results. For this reason, each partner will undertake dissemination and exploitation activities according to their possibilities and profile.
The main results achieved within OD work stream in the 2nd (final) reporting period involve the final system integration of the on-board OD system as well as the finalization and implementation of the software for OD and distance estimation. The functional testing of the fully integrated system was performed in dynamic field tests and the evaluation testing of the integrated OD system was performed in operational environment. The dynamic field tests for the purpose of functional testing as well as the evaluation tests were conducted on the railtrack sites in Serbia, achieving so TRL4-5. The permissions for the tests in operational environment were issued by Serbian Railway upon submission of needed documentations on the developed OD system and its integration onto operational locomotive.
The main results achieved within MY workstream in the 1st reporting period involve analysis and specification of requirements for MY management. Modelling processes in MYs were done and essential data and processes for providing optimisation of available resources and planning of marshalling operations in order to decrease overall transport time and costs with cargo handling were selected and described in details. Standard data formats were analyzed and were taken into account in the initial phase of developing real-time yard management (RTYM) system. Concept solution was done as well as database structure of the RTYM system. Software development for real time decision support system was started based on achieved results in WP4 and WP5. Initial version of Visualization and Planning module with Data Input and Data Output submodules was done during the reporting period.
The main results achieved within MY work stream in the 2nd (final) reporting period involve software development of advisory decision support RTYM system and development of algorithms for processes in local marshalling yard and their optimization. RTYM mainly targets deviations in decision-making processes in marshalling yard with respect to time and infrastructure. This system offers dispatchers and other decision-makers alternative solutions in case of deviations. The main objective is to provide optimization tool to decrease overall transport time and costs with cargo handling. Providing prompt solutions for real-time deviations in decision-making processes with presentation of potential consequences, SMART RTYM increases efficiency on local marshalling yards and decrease time for making decisions. SMART RTYM system was tested on real-time data provided from two local marshalling yards (from Serbia and Bulgaria).
The SMART project results were disseminated at relevant Conferences and other dissemination events. The SMART project results were presented at several relevant external Conferences in the Railway Engineering, Computer Vision and Robotics fields, as well as at the Conferences organised by SMART partners: RAILCON2016, RAILCON2018 and the first International Railway Symposium Aachen IRSA2017. SMART project mid-term results were presented at the Shift Freight to Rail-Midterm Event and the SMART project final event was at the Innovation in Freight Conference, Munich, 18 June 2019.
The SMART obstacle detection demonstrator was displayed at the Shift2Rail stand at the InnoTrans2018, 18th-21st September 2018.
SMART partners will continue keeping on with dissemination of project results beyond the project lifetime to assure better sustainability and usability of the project results. For this reason, each partner will undertake dissemination and exploitation activities according to their possibilities and profile.
The SMART Obstacle Detection System (ODS) advances state of the art in several aspects. SMART developed a novel night vision sensor-camera augmented with image intesifier which was integrated with different vision sensors and a laser sensor. Besides the novel hardware, SMART developed the software that includes novel machine learning-based method for object detection and direct distance estimation able to meet the challenges of railways applications such as long-range obstacle detection. The software development was supported by a newly generated dataset for obstacle detection and distance estimation in railways. In this way, SMART will contribute to long term goal of autonomus cargo haul trains necessary for implementation of EU transport strategy.
The SMART marshalling Yard (MY) management system is a web-based information system for supervision and management of marshalling yards that will enable marshalling yards officials to view marshalling yard representation including receiving yard, hump, classification yard and departure yard. Integration with the marshalling yard automation system should lead to the automation of the marshalling process from planning to execution. SMART MY Management System is a novel machine-learning based decision system that can give optimal or near-optimal solution for any marshalling yard operation in real time.In this way, SMART will contribute to increased efficiency in the MYs and sidings and to reducing lead-time and costs as well as to simplifing of business processes.
The SMART marshalling Yard (MY) management system is a web-based information system for supervision and management of marshalling yards that will enable marshalling yards officials to view marshalling yard representation including receiving yard, hump, classification yard and departure yard. Integration with the marshalling yard automation system should lead to the automation of the marshalling process from planning to execution. SMART MY Management System is a novel machine-learning based decision system that can give optimal or near-optimal solution for any marshalling yard operation in real time.In this way, SMART will contribute to increased efficiency in the MYs and sidings and to reducing lead-time and costs as well as to simplifing of business processes.