SAMIRA
Shunting Assistant & Monitoring Interface for Autonomous Rail Application
Motivation
To combat climate change and an overloaded road infrastructure, more transportation services must be shifted to rail. To achieve this, today's time-consuming and cost-intensive processes and procedures, particularly at rail vehicles’ shunting operations, must be significantly accelerated and efficiency as well as flexibility must be increased tremendously. However, this can only be achieved if the shortage of skilled workers resulting from demographic change in the industry is counteracted, too. SAMIRA can make a key contribution to overcome these challenges and making rail freight transport more efficient and attractive.
The feasibility of this approach has already been demonstrated by the SAMIRA1 project, which was completed from 2019 to 2023 due to a Ziel2-funding program from the state of North Rhine-Westphalia and the European Union.
Objective of SAMIRA2.0
The aim of SAMIRA2.0 is to further develop the system from the previous project and demonstrate its functionality in real-life operations. SAMIRA2 reduces the resources required in the last mile, increases safety in shunting operations and is the basis for autonomous rail operations. It significantly improves the flexibility, cost-effectiveness and competitiveness of rail-transportation and is paving the way for modern, rail-based logistics.
SAMIRA2 is an assistance system for shunting drives in the last mile, whereby modern sensor technology in a portable sensor module (SAMIRAmobil) is attached to the last wagon of a shunting section and thus enables the locomotive driver to carry out shunting drives even without a shunting attendant. For this purpose, the integrated horn and light mounted at the other end of the train can get activated by the engine driver in the driver's cab by the SAMIRA system.
The real-time video stream from the last wagon is shown on a display in the locomotive's driver's cab (SAMIRAhmi), in which important information such as people standing close to the track, remaining distances to obstacles and speed-over-ground are also highlighted by using augmented reality.


In addition, obstacles on the track ahead are detected and, if necessary, a braking recommendation is displayed to the train driver. Important additional information, such as the position of a switch ahead or the predicted braking distance, can also be displayed in the video stream.
This provides a comprehensive overview to the train driver showing what is happening at the other end of the train. The live data is transmitted securely by using 5G mobile radio key technology from the sensor module to the display at the locomotive. The long-term goal is to achieve the ability to carry out shunting movements completely autonomously, like entire CT trains in port areas or on industrial sites.

The photograph shows the SAMIRA demonstrator set up with environmental sensing sensors (LiDAR, radar, cameras), satellite-based positioning (Galileo, GPS), inertial sensor and data transmission of the video stream via a multi-access point mesh network installed on the site.
In SAMIRA2.0, size and weight are now significantly reduced. In addition, now data is transmitted by using the 5G mobile radio system instead of having a private mesh network.
This SAMIRAmobil module also includes integrated image data evaluation to provide augmented reality for identifying obstacles and danger spots, map-based route recognition and prediction of the further course of the journey.
Innovations in the project
The innovations resulting from the project are diverse and form the basis for future autonomous maneuvering. The developments and experiences already made during the earlier construction of the demonstrator form the basis for SAMIRA2.0 and will be further developed and consistently continued. They can be divided into three areas:
AI for dangerous space monitoring in maneuvering operations
Rail operations in the last mile face a particular challenge. In order to enable camera-based object recognition, an artificial intelligence (AI) is being developed and optimized for this application.
- Training of an own developed AI, which is optimized for the special cases existing in the shunting area
- Recognition of signals, signal boards and signs
- Fusion with other sensors for danger zone monitoring

Safe real-time data transmission
Very strict safety and real-time requirements apply to data transmission from the sensor modules into the driver's cab of the locomotive. Based on the experiences made in the past during the construction of the demonstration system, data transmission in SAMIRA2.0 is no longer realized by a proprietary multi-mesh WLAN system, but it is implemented on the basis of a 5G network, which provides.
- Secure and latency-minimized real-time data transmission from/to the moving train via the public 5G network
- Use of private campus networks to compensate dead spots and to provide communication within factory halls
- Seamless handover between public 5G network and local 5G campus network during operation
High precise positioning
A highly accurate, precise and track-selective position calculation is essential for a safety system. The track-selective positioning solution that has already been developed delivered very good results in the demonstrator already. In addition, now the lateral positioning accuracy on the track is being further improved.
- Use of a GNSS-INS system with RTK to achieve precision in the centimeter range
- Use of digital maps of the tracks for stabilizing the position solution

SAMIRA is groundbreaking
SAMIRA is a driver assistance system for freight trains in bidirectional operations, consisting of comprehensive environmental sensors on the train combined with artificial intelligence (AI). In association with a highly secured state-of-the-art wireless data transmission in real time, the area surrounding a freight train and its routes can be monitored and controlled safely. This data can be used to provide the train engineer with comprehensive information and, in the event of an incident, for immediate warnings. The sensors and their analysis with AI are supporting people and greatly increase safety.
Since the augmented video stream is transmitted via a data network, this data is available to anyone who has proper authorization to connect to the net, e.g. employees in the control center or people working on the track or the operations’ supervisors.
In the long term, SAMIRA enables digitalization in the last mile, reduce the resources required and form an initial basis for autonomous rail operations, e.g. if a SAMIRAmobil module is attached to each of both ends of the train or if a fixed installation is implemented in the shunting locomotive. Nevertheless, this will significantly improve the competitiveness of rail as a mode of transport.

Another forward-looking feature is secure data transmission using 5G technology. With these increasing and widely available networks, any data can be transmitted, e.g. technical operation data to a control center or company communication radio or any data supporting Logistics 4.0.
The core components of the SAMIRAmobil module can also be configured as a stationary module (SAMIRAfix). This module can be permanently installed on a pole or at a building sensing and watching hazardous hot spots such as railway crossings or complex and confusing track ladders, for example. The integrated sensors provide reliable data and hazard warnings as well as an overview of local conditions via a real-time live stream.
SAMIRA is a modular, scalable system that can be tailored to all conditions in the area and the respective requirements. Also, it can be expanded and adapted at any time in the future. The standardized 5G technology as backbone network ensures extremely secure and deterministic real-time data transmission in the network by using standard industrial components, which can also be repaired very quickly in the event of a failure.
The SAMIRA2.0 project will run until March 2026. Before the project will get closed, there will be a real live-demonstration on an industrial shunting yard showing the entire systems and its results. If you are interested in our work, please feel free to contact us - we would be more than happy to hear from you!
Acknowledgments:The entire project team would like to expressly thank the German Federal Railway Authority (Eisenbahn-Bundesamt EBA) and the German Centre for Rail Transport Research (Deutsches Zentrum für Schienenverkehrsforschung DZSF) for the approval of the project and the funding received for the SAMIRA2.0 research project. Without this support, the project could not be realized. Förderkennzeichen: BMDV 53T23001IK.