Proceedings of the Institution of Mechanical Engineers,Part F: Journal of Rail and Rapid Transit

The maintenance of bogie components, a critical aspect of railway maintenance, is difficult due to the confined underframe space. This makes it difficult to install traditional monitoring equipment, resulting in a labour-intensive process. Thus, a lot of time has to be expended to conduct these tests, which makes the process both tedious and expensive. Moreover, this approach is somewhat inadequate, since the tests can only be conducted at the depot and thus only when the trains are out of service. We have developed and deployed a non-intrusive solution based on a small wireless sensor network that can be easily installed on the different parts of the bogie and along the whole train. We have worked out a technique to discriminate between the various sources of vibration and can thus monitor the state of several components using only a few sensors. In this paper, we present a case study on how to maintain an axle-box and a wheel-set by attaching a single intelligent sensor to the bogie frame or the bearing cover and using the empirical mode decomposition technique to analyse the generated data. In light of the promising results obtained in this study, we suggest that the proposed approach can lead to a value-added predictive maintenance strategy as long as the test conditions are kept under control. However, we do highlight that the generalization of the approach relies on the flexibility of the system to adapt to new environments and operational scenarios.

Alarm management is a key component of the successful operation of a prognostic or health-monitoring technology. Although alarms can alert the operator to critical information, false alarms and alarm flooding can cause major difficulties for successfully diagnosing and acting upon infrastructure faults. Human factors approaches seek to design more-effective alarm systems through a deep understanding of the contextual factors that influence alarm response, including strategies and heuristics used by operators. This paper presents an extensive analysis of alarm-handling activity in the setting of an Electrical Control Room on the rail network. The analysis is based on contextual observation, and the application of a time-stamped observation checklist. Functions, performance requirements, and general operating conditions that influence alarm handling are presented, delineating the typical operational constraints that need to be considered in the design and deployment of asset-based alarm systems. The analysis of specific alarm-handling incidents reveals the use of specific strategies that may bias operator performance. Implications for the design of health-monitoring systems are discussed.

In the financial year 2009/10, the Great Britain (GB) rail infrastructure manager, Network Rail, spent £32 million on the failures within switches and crossings. Approximately 53% of those failures occurred within the switch panel. In addition, two major incidents in GB in recent years have highlighted a lack of understanding of the loads and vibrations experienced by, and the consequent rates of deterioration of, switch panels. This paper describes work which has been undertaken to help improve understanding of in-service loads experienced by switch panels and their consequent deterioration rate. Field experimentation has been designed and installed on four sites around GB, with the same design of switch. The change in response to loading, and the rate of deterioration, of the switch panels at each site was monitored over time. The effect of the vehicles and the deterioration were analysed individually before a comparison between the four sites. The analysis from the strain gauge measurements showed that there was an increase in the variance and maximum strains generated on the stock rail with the switch closed compared to when the switch is open. Finite element analysis was used to validate variation recorded by the strain gauges under similar loads.

This paper presents qualitative and quantitative analyses of the action of points, a critical component of railway networks. They allow diagnostic and prognostic evaluations of the points using health monitoring systems, i.e. they allow the state of the system at a desired moment to be evaluated and the forecasting of the future condition of the system. The main objective is to increase the reliability, availability, safety and maintainability of these systems. A novel approach for maintenance management based on fault tree analysis is proposed. A binary decision diagram (BDD) approach is proposed for the qualitative analysis of the fault trees. The BDD obtains a Boolean expression for the fault tree. An optimal ordering of events is required in order to obtain an efficient conversion from a fault tree to a BDD, with the AND method being used for this purpose. Each event is classified based on its importance to the fault tree. It is studied using the Birnbaum and Criticality importance measures, using the Boolean expressions in order to perform accurate diagnostics and valuable prognostics on the state of the system. The presented approach allows the failure probability of the system to be determined along with importance measures obtained by considering variable time increments, e.g. shorter periods at the beginning and end of the life cycle of an event. A real case study on a set of M63 points is presented. The results provide useful information that can be used to support operations and the planning of maintenance tasks. The approach creates a methodology to establish effective maintenance planning, as it is a flexible and simple method that takes into account a nonlinear system that leads to an optimal allocation of resources. Finally, the conditions for an optimized online decision-making process are achieved.

In reliability and maintenance engineering, availability can be described as the ability of an item to be in a state to perform a required function at a given time. Availability is commonly given as a measure between zero and one, where one means the probability of an item to be available for use at a given time is 100%. Availability is measured in many areas, such as electronics, information technologies, military equipment, electrical grids and the industry. Various indicators related to availability of railways have been examined by academia and industry. However, there is some ambiguity about how to define and measure the availability of rail infrastructure, given railways' semi-continuous operation, besides data quality issues. This article considers the application of common definitions of availability to rail infrastructure. It includes a case study comparing various approaches for measuring availability. The case study ends with a section on how availability as a function of train frequency and maintenance time can be simulated. The results show rail infrastructure availability correlates well with train delay, but this depends on how infrastructure failure data and outliers are treated.

Railway assets suffer wear and tear during operation. Prognostics can be used to assess the current health of a system and predict its remaining life, based on features that capture the gradual degradation of its operational capabilities. Prognostics are critical to improve safety, plan successful work, schedule maintenance, and reduce maintenance costs and down time. Prognosis is a relatively new area; however, it has become an important part of condition-based maintenance of systems.

As there are many prognostic techniques, usage must be tuned to particular applications. Broadly stated, prognostic methods are either data driven, or rule or model based. Each approach has advantages and disadvantages, depending on the hierarchical level of the analysed item; consequently, they are often combined in hybrid applications. A hybrid model can combine some or all model types; thus, more-complete information can be gathered, leading to more-accurate recognition of the impending fault state. However, the amount of information collected from disparate data sources is increasing exponentially and has different natures and granularity; therefore, there is a real need for context engines to establish meaningful data links for further exploration.

This approach is especially relevant in railway systems where the maintainer and operator know some of the failure mechanisms, but the sheer complexity of the infrastructure and rolling stock precludes the development of a complete model-based approach. Hybrid models are extremely useful for accurately estimating the remaining useful life (RUL) of railway systems. This paper addresses the process of data aggregation into a contextual awareness hybrid model to obtain RUL values within logical confidence intervals so that the life cycle of railway assets can be managed and optimized.