Russia: The R&D Institute of Russian Railways (RZD) is integrating diagnostic telematics devices in 9 ED4M trains at the depot of the Moscow Railway. Also, the project of the VNIIZhT computer vision system for track machines reached the final of the New Link competition.
The onboard device for the automated multiple unit trains operation control system (BU AS KR MVPS) was developed. It allows to check a number of parameters (door position, cabin temperature, speed, train location) and display this information on a monitor in the driver’s cab. For example, the geolocation data helps to control whether the train follows the schedule. As regards the doors operation, the technology should allow avoiding the grasping of passengers, reduce the loss of time to determine a specific open door, etc.
(enlarge). Source: VNIIZhT
As Veronika Fedorova, Head of the Integrated Innovation Projects Department in the Data-Analysis Research Center of VNIIZhT, told the Gudok newspaper, the system has already been installed and is being commissioned on 9 eleven-car ED4M EMUs at one of the depots in Moscow. The technology should become the basis for the unified commuter trains diagnostics & monitoring system called AS DMPS, which should aggregate the data from the train on-board devices.
VNIIZHT obtained a patent for AS DMPS in June. It also declares that the system will also monitor the car occupancy, the level of lightning inside the train, humidity and the percentage of CO2, dust and other substances in the air. For that to happen, the cars will be equipped with autonomous wireless control sensors. Their number may vary depending on the requirements of the operator. The GLONASS/GPS navigation modules will be installed in the head cars of a train. The occupancy of the cars is planned to be monitored by 4 sensors installed on axle boxes of bogies. They will measure the weight of the car in real-time: the weight of the empty car will be automatically subtracted from this figure, and then the resulting amount will be divided by 80 kg (the average weight of one passenger with luggage when traveling on a commuter train).
(enlarge). Source: VNIIZhT
The main page of the AS DMPS interface (enlarge). Source: VNIIZhT
It is planned that the data from the sensors via cellular communication channels will flock to the data processing server, from where the sorted information divided by cars will be transmitted to the mobile applications both for staff and passengers. It is assumed that they will be able to send messages to the monitoring center about the discrepancy between the actual parameters of the service quality displayed in the application. Currently, there are more than 11,000 cars of ET2M, ED2T and ED4M trains on the RZD’s network.
Also, the VNIIZhT’s project of computer vision for track machines’ operational readiness reached the final of the New Link, the youth competition of the Russian Railways. The VNIIZhT team has been working on it for more than a year and a half.
This project is intended to automate the control of the readiness of track sections for straightening and is supposed to be integrated into the previously developed technology by VNIIZhT, the automated system for track machines operation monitoring (AS KRSPS SADKO). The last one is already widespread on the RZD network, and the computer vision system is planned to supplement it with the options to automatically determine the filling of sleeper boxes with crushed stone, the presence and condition of rail fasteners, the critical state of the sleepers, and etc. In this case, the final decision on the admission of straightening vehicle will be made by the head of such works taking into account the conclusion of the system. As the authors of the project note, the operation of this system does not depend on weather conditions or time of day, it is mounted without changes to the track machine construction. The system has already passed preliminary tests at Oktyabrskaya Railway, and the results confirmed the designed performance.
About the AS KRSPS SADKO operation. Source: VNIIZhT, youtube.com
According to Veronika Fedorova, it is proposed to install cameras, lidars, lighting devices and a computing module with AI on track machines within the framework of the project. “The deep learning process took a lot of time, it was necessary to take into account all the options for sections of the track as much as possible and scan and analyze them”, she notes. The equipping of the RZD main fleet of straightening and snow-removing vehicles with the AS KRSPS system is almost completed. More than 3,500 track machines of over 100 types and modifications are connected to it.