Five steps to achie­ving fully au­to­no­mous track main­te­nance machines

The path to fully autonomous track maintenance machines leads through five levels of automation: from a manual work process to assisted and semi-autonomous working to fully autonomous working and running. AI and sensor technology take care of analysis, control, and documentation. This reduces staff requirements and increases safety.

The field of track construction and maintenance is undergoing profound changes. A shortage of skilled labour, strict quality requirements, and increasing economic pressure call for new solutions. Plasser & Theurer has responded by clearly defining the path for developing fully autonomous track maintenance machines. It involves a systematic evolution defined by five levels of automation.

At the core, there is the idea that the machine takes over tasks that require experience, full focus, and a high number of staff today. This increases precision with repetition and eliminates sources of error. The machine becomes a reliable partner that reduces the staff’s workload while increasing safety on the worksite.

At the same time, the role of the operating staff is changing: the focus is less on manual skills as system understanding and monitoring expertise are gaining importance. Networked machines, digital measuring processes, and centralized data hubs are creating a new form of maintenance that relies on predictive planning instead of reactive repairs.

Level 1 - Manual work process: humans control everything

Operating staff control all maintenance tasks themselves. Experienced specialists are responsible for every function of the machine, from positioning and process control to quality assurance. Mechanical or simply digitally based input devices are used to operate the machine. Precision with repetition and efficiency depend directly on the skills and experience of the employees.

Level 2 - Assisted work process: digital helpers in the background

Assistance systems support the operating staff in individual sub-processes. The machine makes recurring tasks easier, improves ergonomics and precision with repetition, but does not yet take over process control independently. One example of this is TampingControl. The system monitors the tamping parameters and adjusts them automatically. Quality, monitoring, and responsibility remain entirely with the staff. Level 2 marks the entry into digital process support without replacing human control.

Level 3 - Semi-autonomous work process: first maintenance tasks run independently

The machine begins to play an active part in the work process. It carries out individual maintenance tasks independently, e.g. tamping, lifting, or lining. However, this is done within clearly defined framework conditions. The operating staff monitor the processes and safeguard them by intervening in the event of deviations or failures. Requirements are shifting: controlling processes manually becomes less relevant compared to system understanding, handling sensor technology, and digital control processes. Working and driving cabs remain in place because operators are still present on-site.

Level 4 - Fully autonomous work process: the machine controls the entire process

The machine takes care of all maintenance tasks independently – from analysing the environmental conditions to process execution and quality assurance. It reacts independently to deviations within a defined scope of action and carries out the work tasks completely autonomously. The operating staff are no longer involved in the process and only intervene manually in exceptional cases.  As they do not need to actively control the process, a work cab is no longer necessary. Only one driving cab remains for transfer and monitoring, because the driving functions have not been automated yet.

Level 5 - Fully autonomous machine: travelling completely independently

The track maintenance machine takes over all driving and work processes independently. This includes quality checks and troubleshooting - without human intervention on-site. Responsibility for the system lies with the manufacturer, while control and monitoring are carried out exclusively remotely, for example via a central control or operations centre. Working and driving cabs are no longer required, as no staff are needed on the machine for either the working process or transfer travel. This maximizes safety: no employees in the danger zone, no operating errors, and consistently high precision with repetition.

Benefits throughout the five levels

Operation and operating staff. Each automation step reduces the operating effort. At levels 1 and 2, quality depends heavily on the operators’ level of experience. Assistance systems such as TampingControl reduce the workload and standardize some subtasks. From level 3, the machine takes over core maintenance tasks independently, while staff is responsible for monitoring and managing the system. At level 4, only one person is needed at the control desk, as the machine controls the entire work process autonomously. At level 5, the machine is controlled entirely from a central control centre. Neither working nor driving processes need to be operated on-site. This reduces staff requirements, facilitates deployment planning, and makes track construction and maintenance more resilient to shortages of skilled labour.

Process quality. The higher the level of automation, the more process knowledge is transferred to the machine. Initially, operating staff must take tamping parameters, standards, and specifications into account themselves. From level 3, the machine generates these parameters automatically based on sensor data and digital sets of rules. At level 4, it also takes over quality assurance completely. Systems such as TampingReport document every step digitally and make the results transparent and traceable, regardless of who operates the machine. This results in stable, reproducible work quality. This, in turn, reduces sources of error and minimizes subsequent corrections.

Safety and cost-efficiency. Increasing autonomy reduces the risk for staff, as fewer and fewer people have to work in the danger zone. Assistance systems recognize obstacles, monitor processes, and prevent operating errors. From level 4, the machine works largely without a human present at the worksite. At Level 5, working and driving processes are carried out completely autonomously. At the same time, downtimes are reduced thanks to continuous data analysis, predictive maintenance, and faster error diagnostics. Networked machines optimize material usage and work processes, e.g. by transferring data in real time between a tamping machine and a ballast profiling machine. This lowers costs along the entire life cycle and increases the availability of the infrastructure.

Autonomy becomes reality

Elements of the individual automation stages are already installed in machines today and are proving their practical suitability under real conditions. This makes it clear that autonomous functions have become a part of everyday life. They provide concrete added value and show where railway construction is heading in the future.