The use of new technologies in the field of sensor technology and the targeted application of data analysis methods is intended to make plant maintenance more efficient and cost-effective. Ideally, the maintenance interval is then selected so that an incipient fault can be detected, but this has not yet led to a failure. Predictive condition-based maintenance is intended to avoid both unnecessary maintenance work and failures.
Although the general goals and associated challenges for infrastructure operators are similar within a layer, it is not possible to develop a universal planning tool that delivers optimal results for all users without customization. This is due to the diversity of data structures and company-specific planning constraints.
As part of the Shift2Rail project IN2SMART2, we have developed specific planning tools in close cooperation with our industry partners. These tools enable predictive maintenance planning at different planning levels, based on forecasted condition developments and formalized expert knowledge.
Level: strategic (based on simulated tactical planning)
Task: Support strategic planning decisions by simulating and evaluating the tactical planning process over a rolling time horizon in different scenarios.
Level: operational
Task: Develop a lean tamping plan based on the predicted condition of the track sections so that a specified network quality is ensured, and the available budget is met.
Level: operational
Task: Preparation of a shift plan for tamping activities based on the condition prediction for 25 m long sections.
To ensure the transferability of the planning tools to other infrastructure systems, attention was paid to a modular solution approach, which can be adapted to different user-specific data and system requirements.
Due to the constantly increasing demand in passenger and freight transport, capacity bottlenecks are becoming more and more frequent. It is therefore essential to make infrastructure maintenance as effective as possible. The development of complex measuring systems enables continuous monitoring of the condition of the tracks, the track bed or the switches. This reduces the effort required for manual inspections.
In a European research initiative, various research institutes and railway maintenance companies are working together to develop an automated, condition-based and real-time maintenance management system (IAMS) for rail infrastructure. In this framework, analysis and prediction algorithms are used to provide information on the current and forecasted status of the rail infrastructure. These predictions are always affected by uncertainties and are taken into account in the decision-making process as probabilistic information. Fraunhofer IVI develops robust optimization algorithms to simplify and automate the complex, multi-stage decision-making process of maintenance planning.
To simplify the planning process, planning in maintenance management is generally divided into three levels: strategic, tactical and operational. The strategic planning level covers the longest time horizon. Here, basic strategies for maintenance, threshold values as well as maintenance technologies used are defined. The second level, that of tactical planning, comprises medium-term planning that considers complex interventions with long lead times. The most detailed planning, which covers the shortest time horizon, is the lowest operational planning level. Here, detailed operational planning is prepared, taking into account capacity bottlenecks, in order to ensure that all interventions can be executed smoothly. In addition, it is necessary to be able to react dynamically to unexpected and unforeseen failures or changes during the execution of the plans.
Innovative technologies for rail transport
The Shift2Rail Joint Technology Initiative (JTI) aims to improve the quality and efficiency of railway services through the incorporation of innovative technologies. To achieve this goal, new technologies that comply with the highest safety standards are to be developed, implemented, demonstrated and validated.
Fraunhofer Institute for Transportation and Infrastructure Systems IVI