Adithya Thaduri

Railway Assets: A Potential Domain for Big Data Analytics☆

Abstract Two concepts currently at the leading edge of todays information technology revolution are Analytics and Big Data. The public transportation industry has been at the forefront in utilizing and implementing Analytics and Big Data, from ridership forecasting to transit operations Rail transit systems have been especially involved with these IT concepts, and tend to be especially amenable to the advantages of Analytics and Big Data because they are generally closed systems that involve sophisticated processing of large volumes of data. The more that public transportation professionals and decision makers understand the role of Analytics and Big Data in their industry in perspective, the more effectively they will be able to utilize its promise. This paper gives an overview of Big Data technologies in context of transportation with specific to Railways. This paper also gives an insight on how the existing data modules from the transport authority combines Big Data and how can be incorporated in providing maintenance decision making.

Computational intelligence framework for context-aware decision making

Learning of context-aware systems is necessary in building up knowledge on the characteristics of the environment to provide efficient decision making within multi-objective requirements. As the industrial systems becomes complex day-by-day, intelligent machine learning techniques need to be implemented at respective context-aware situations to facilitate recommendations using soft computing methods based on dynamic user specifications. In this paper, a framework is designed for a meta-database that is generated by contextual information of several peers with what-if conditions and rule-based approaches and thus by provide decision making utilizing several existing soft computing algorithms.

Maintenance 4.0 in Railway Transportation Industry

Transportation systems are complex with respect to technology and operations with involvement in a wide range of human actors, organisations and technical solutions. For the operations and control of such complex environments, a viable solution is to apply intelligent computerised systems, such as computerised traffic control systems for coordinating airline transportation, or advanced monitoring and diagnostic systems in vehicles. Moreover, transportation assets cannot compromise the safety of the passengers by applying operation and maintenance activities. Indeed safety becomes a more difficult goal to achieve using traditional maintenance strategies and computerised solutions come into the picture as the only option to deal with complex systems interacting among them trying to balance the growth in technical complexity together with stable and acceptable dependability indexes. Industry 4.0 is a term that describes the fourth generation of industrial activity which is enabled by smart systems and Internet-based solutions. Two of the characteristic features of Industry 4.0 are computerization by utilising cyber-physical systems and intelligent factories that are based on the concept of “internet of things”. Maintenance is one of the application areas, referred to as maintenance 4.0, in form of self-learning and smart systems that predicts failure, makes diagnosis and triggers maintenance by making use of “internet of things”. This paper discusses the possibilities that lie within applying the maintenance 4.0 concept in the railway transportation industry and the positive effects on technology, organisation and operations from a systems perspective.

Comparison of reliability prediction methods using life cycle cost analysis

In this paper, it was discussed on the several reliability prediction models for electronic components and comparison of these methods was also illustrated. A combined methodology for comparing the cost incurring for prediction was designed and implemented with an instrumentation amplifier and a BJT transistor. By using the physics of failure approach, the dominant stress parameters were selected on basis of research study and were subjected to both instrumentation amplifier and BJT transistor. The procedure was implemented using the methodology specified in this paper and modeled the performance parameters accordingly. From the prescribed failure criteria, mean time to failure was calculated for both the components. Similarly, using 217 plus reliability prediction book, MTTF was also calculated and compared with the prediction using physics of failure. Then, the costing implications of both the components were discussed and compared them. From the results, it was concluded that for critical components like instrumentation amplifier though the initial cost of physics of failure prediction is too high, the total cost incurred including the penalty costs were lower than that of traditional reliability prediction method. But for non-critical components like BJT transistor, the total cost of physics of failure approach was too higher than traditional approach and hence traditional approach was much efficient. Several other factors were also compared for both reliability prediction methods.

Investigation of Causes of Mining Machines Maintenance Problems

Human errors in the area of mining engineering are of critical issue that has serious concerns in safety, operation and production performance. There is a need for finding cause and effect relations with respect to the maintenance issues in order to detect, scrutinize and take necessary actions to reduce it. This paper deals with the human errors in the mining machines for the maintenance problems using fishbone cause and effect analysis. The investigation of these causes and effects are carried out during different operating conditions in typical mining industry and potential problems are assessed. There are several recommendations are provided to reduce the effect of human error so as to increase production by careful consideration of maintenance activities.

Reliability prediction of constant fraction discriminator using modified PoF approach

In this paper, the introduction, functioning and importance of constant fraction discriminator in nuclear field was studied. Furthermore, reliability and degradation mechanisms that affects the performance of output pulse with temperature and dose rates acts as input characteristics was properly explained and verified with the experiments. Accelerated testing was carried out to define the life testing of the component with respect to degradation in output TTL pulse amplitude. Time to failure was to be properly quantified and modelled accordingly.

Comparison of failure characteristics of different electronic technologies by using modified physics-of-failure approach

The electronic components are used in several safety and maintenance systems that require accurate reliability prediction for higher availability. The traditional reliability prediction methods that draw on standard handbooks such as MIL-HDBK 217F, Telcordia, CNET etc., are inappropriate to determine the reliability indices of these components due to empirical methods does not comply with operating life cycle and technology advancements. The progressive reliability prediction methodology, the physics-of-failure (PoF), emphasizes the root cause of failure, failure analysis, and failure mechanisms based on the analysis of parameter characteristics. However, there is a limitation: it is sometimes difficult to obtain manufacturer’s details for failure analysis and quality information. Several statistical and probability modeling methods can be performed on the experimental data of these components to measure the time to failure. These experiments can be conducted using the accelerated-testing of dominant stress parameters such as voltage, current, temperature, radiation etc. In this paper, the combination of qualitative data from PoF approach and quantitative data from the statistical analysis is used to create a modified physics-of-failure approach. The critical electronic components used in certain safety systems from different technologies are chosen for reliability prediction: optocoupler, constant fraction discriminator, BJT transistor, voltage comparator, voltage follower and instrumentation amplifier is studied. The failure characteristics of each of the technologies are studied and compared according to operating conditions.

FAILURE MODELING OF CONSTANT FRACTION DISCRIMINATOR USING PHYSICS OF FAILURE APPROACH

Due to several advancements in the technology trends in electronics, the reliability prediction by the constant failure methods and standards no longer provide accurate time to failure. The physics of failure methodology provides a detailed insight on the operation, failure point location and causes of failure for old, existing and newly developed components with consideration of failure mechanisms. Since safety is a major criteria for the nuclear industries, the failure modeling of advanced custom made critical components that exists on signal conditioning module are need to be studied with higher confidence. One of the components, constant fraction discriminator, is the critical part at which the failure phenomenon and modeling by regression is studied in this paper using physics of failure methodology.

TWO-STAGE DESIGN OF EXPERIMENTS APPROACH FOR PREDICTION OF RELIABILITY OF OPTOCOUPLERS

Conventionally, reliability prediction of electronic components is carried out using standard handbooks such as MIL STD 217 plus, Telcordia, etc. But these methods fail to provide a realistic estimate of reliability for upcoming technologies. Currently, electronic reliability prediction is moving towards applying the Physics of Failure approach which considers information on process, technology, fabrication techniques, materials used, etc. Industries employ different technologies like CMOS, BJT and BICMOS for various applications. The possibility of chance of failure at interdependencies of materials, processes, and characteristics under operating conditions is the major concern which affects the performance of the devices. They are characterized by several failure mechanisms at various stages such as wafer level, interconnection, etc. For this, the dominant failure mechanisms and stress parameters needs to be identified. Optocouplers are used in input protection of several instrumentation systems providing safety under over-stress conditions. Hence, there is a need to study the reliability and safety aspects of optocouplers. Design of experiments is an efficient and prominent methodology for finding the reliability of the item, as the experiment provides a proof for the hypothesis under consideration. One of the important techniques involved is Taguchi method which is employed for finding the prominent failure mechanisms in semiconductor devices. By physics of failure approach, the factors that are affecting the performance on both environmental and electrical parameters with stress levels for optocouplers are identified. By constructing a 2-stage Taguchi array with these parameters where output parameters decides the effect of top two dominant failure mechanisms and their extent of chance of failure can be predicted. This analysis helps us in making the appropriate modifications considering both the failure mechanisms for the reliability growth of these devices. This paper highlights the application of design of experiments for finding the dominant failure mechanisms towards using physics of failure approach in electronic reliability prediction of optocouplers for application of instrumentation.

Process Mining for Maintenance Decision Support

In carrying out maintenance actions, there are several processes running simultaneously among different assets, stakeholders, and resources. Due to the complexity of maintenance process in general, there will be several bottlenecks for carrying out actions that lead to reduction in maintenance efficiency, increase in unnecessary costs and a hindrance to operations. One of the tools that is emerging to solve the above issues is the use Process Mining tools and models. Process mining is attaining significance for solving specific problems related to process such as classification, clustering, discovery of process, prediction of bottlenecks, developing of process workflow, etc. The main objective of this paper is to utilize the concept of process mining to map and comprehend a set of maintenance reports mainly repair or replacement from some lines on the Swedish railway network. To attain the above objective, the reports were processed to extract out time related maintenance parameters such as administrative, logistic and repair times. Bottlenecks are identified in the maintenance process and this information will be useful for maintenance service providers, infrastructure managers, asset owners and other stakeholders for improvement and maintenance effectiveness.