The driver assistance system for shunting freight trains

Shunting Assistant & Monitoring Interface
for Autonomous Rail Applications (SAMIRA)

Shunting Today

1.   The Initial Situation

SAMIRA was triggered by today’s time and cost-intensive processes and procedures in the shunting operation of rail vehicles, e.g. in port logistics or in multimodal transhipment hubs. The very tense personnel situation in transport and logistics companies even multiplies the problem.

Today, shunting operations with long freight trains are practised in two-man operations since one-man operation using radio control system is far to time-consuming and inefficient due to the associated walking times when changing the direction. In addition, the use of radio remote controls for trains longer than 550 m in certain areas is not a safe solution due to interferences on radio transmissions. For short shunting distances, however, radio remote control is the preferred operation method, though.

In the long-term, the goal is to carry out shunting operations autonomously, e.g. entire freight trains in port areas. In the short term, even the reduction of the required human resources per train will help, e.g. for being able to carry out the shunting operations with only one engine driver. For this purpose, technical and organisational solutions must be developed which enable the locomotive driver to have a reliable overview of the area where he is driving along. A dedicated sensor system developed for this purpose provide him a seamless, safe and highly reliable monitoring of the track area during pulled and pushed shunting movements. In a second stage, this technology can even be the cornerstone for enabling automated shunting operations in limited areas, i.e. as a preliminary stage to autonomous operation.

2.   The SAMIRA Solution

This situation initiated the development of the SAMIRA concept depicted below, in which both ends of the train are equipped with a camera and other environmental sensors (e.g. RADAR, LiDAR). These sensor modules, combined with a very safe and low-latency real-time transmission of the live image captured by a camera and the sensor data to the driver’s cab, enables the driver to safely perform a pushed ride on sight without having to leave the locomotive. Both, the sensor technology for monitoring the driving space and the data transmission are safety-relevant components that must fully comply with the railway-specific requirements.

Structure and building blocks of SAMIRA concept

SAMIRA mobile mounted a the end of the last wagon (principal graphic)

In addition to the real time video stream, the train driver is provided with a variety of information, data and warnings projected into the video stream.

Additional stationary modules (SAMIRAfixed) for object detection on safety critical areas (level crossings, e.g.) increase safety on the track while an RTK station for precise position determination enables the train position to be located precisely on the track.

The data obtained will be used to create a live map of the transfer- and freight-stations. This will further increase safety, enable greater flexibility in shunting operations and be available to all employees with mobile devices that have been registered and connected to the network.

Example: camera image with augmented reality and display of the stopping distance
Concept digital map
Live map of the railway premises with danger spots highlights

3.   Innovations within the Project

The project’s innovations are manifold and together they form the basis for future autonomous manoeuvring. They can be grouped into three areas:

Danger area monitoring in shunting operation:

  • SAMIRA comprises a portable sensor unit (SAMIRAmobile wagon) attached at the end of the last freight wagon for monitoring the danger zone during pushed shunting. By using radar and lidar sensors as well as cameras, the locomotive driver at the other end of the train receives a secured live image of the danger zone in front of the heading waggon.
  • In addition, the locomotive driver receives information about the objects and their sizes and distances, vehicle speed, required braking distance, etc., as an overlay onto the live image – i.e. Augmented Reality.
  • SAMIRA replaces the shunting attendant’s view and voice radio, which in turn allows to use the attendant as additional locomotive driver for another train. SAMIRA helps to alleviate today’s staff shortage.
  • If such a sensor unit is also attached in front to the locomotive (SAMIRAmobile locomotive), the system supports the driver even when driving towards pulling direction. The predictive sensor system also helps to avoid accidents in bad visibility conditions.
  • SAMIRA detects dangerous situations and provides audio-visual warnings.
  • By adding additional sensor modules (SAMIRAfix), which are stationary fixed at dedicated danger spots, a schematic live map of the entire railway area can be created by superposition of all senor modules. Hence, the tracks, trains and other objects on the tracks are displayed in real time.
  • The high-precision position determination used in SAMIRA in conjunction with radar and camera sensors allows time- and energy-optimized braking and acceleration processes that are being calculated and displayed to the train driver via an assistance program. This will also greatly reduce squeaking noises during braking and noises impacted by shunting.
  • If the control signals from SAMIRA are actively connected to the locomotive’s control electronics, they can also be used for automatic braking on target position (this feature not part of the project scope due to lengthy certification and approval processes).

Data transmission via a multipath mesh network

  • In order to meet the strict real-time and safety requirements for data transmission from the sensor modules to the locomotive driver’s cab, a real-time and roaming-capable, professional multipath mesh network including the necessary protocols is being explicitly developed within the project.
  • In addition, SAMIRA provides an interface for online transmission of locomotive operating data to the control centre. The data can be evaluated at a later stage in order to develop further accident prevention in the long term.
  • The infrastructure can also be used as a basis for a higher level of logistics’ automation (Logistic 4.0), e.g. by automatically issuing and transmitting transport orders or shunting plans.

Highly accurate positioning solution

  • Since the position accuracy of a simple SatNav receiver is not sufficient for fulfilling the requirements in the shunting area, different methods are being used to obtain a precise position, compared and evaluated for their suitability during this research project.
  • Based on the European Navigation Satellite System (EGNSS), additional sensors (IMU, LiDAR) and physical measurement variables (odometer) are included into the positioning solutions and further optimized by means of machine learning.

4.   SAMIRA is paves the way to the future of shunting

In a first step, SAMIRA is an advanced driver assistance system (ADAS) for freight trains in bi-directional shunting operations. It consists of comprehensive sensors combined with artificial intelligence (AI) to detect and monitor the train’s environment and dangerous spots. In conjunction with secure, wireless data transmission in real time, the environment of a freight train and its tracks can be monitored safely and the train driver can be informed comprehensively and, in the event of an incident, warned immediately. The sensors and their evaluations with AI support humans and increase safety.

Far beyond that, the SAMIRA system will enable digitalization in the last mile of transportation, reduce the required resources and form an initial basis for autonomous railway operations. This will significantly improve the competitiveness of rail as a mode of transport.

The leading-edge mesh network, which already provides the basic structure for secure and low-latency data communication for Logistics 4.0, is also forward-leading since new additional devices can be flexibly and quickly integrated.

The SAMIRA project will run until August 2022, and before its end there will be a real demonstration of the entire SAMIRA system on the RheinCargo site.

If you are interested in our work, please feel free to contact us, we would be very pleased!

All graphics: SAMIRA project team


The entire project team would like to expressively thank the state of North Rhine-Westphalia and the European Commission for the funding received for the SAMIRA research project, otherwise the project could not have been realised.