Rengarajan Srinivasan

Value of condition monitoring in infrastructure maintenance

As many civil infrastructure assets age over time and often exceed their design life, there is an increasing need to assess and monitor their condition so as to allow improved and effective maintenance decisions. In this regard, there is growing interest in implementing advanced condition monitoring techniques to capture the condition of the infrastructure assets. However, there is a lack of understanding in determining the benefits offered by the various condition monitoring techniques, such that the most appropriate system is implemented to reap the benefits. In this context, this paper proposes a methodology to assess the benefits offered by condition monitoring systems and the factors that affect the value delivered by such systems. The proposed methodology is based on Partially Observable Markov Decision Process (POMDP) and is illustrated with an industrial case example.

Semi-Markov Decision Process With Partial Information for Maintenance Decisions

A critical factor that prevents optimal scheduling of maintenance interventions is the uncertainty regarding the current condition of the asset under consideration, as well as the rate at which deterioration takes place. However, current maintenance modeling and optimization techniques assume that the condition of the asset is either known, or assumed to have an exponential deterioration rate. In this paper, we present a novel approach to maintenance modeling that removes such assumptions. Here, we employ a Partially Observable Semi-Markov Decision Process (POSMDP) for optimizing maintenance decisions, where the condition of the asset is not fully observable, and decision epochs occur at times following any other type of distribution. This method enables a more realistic way of modeling asset deterioration and optimizing maintenance schedules.

On fault propagation in deterioration of multi-component systems

In extant literature, deterioration dependence among components can be modelled as inherent dependence and induced dependence. We find that the two types of dependence may co-exist and interact with each other in one multi-component system. We refer to this phenomenon as fault propagation. In practice, a fault induced by the malfunction of a non-critical component may further propagate through the dependence amongst critical components. Such fault propagation scenario happens in industrial assets or systems (bridge deck, and heat exchanging system). In this paper, a multi-layered vector-valued continuous-time Markov chain is developed to capture the characteristics of fault propagation. To obtain the mathematical tractability, we derive a partitioning rule to aggregate states with the same characteristics while keeping the overall aging behaviour of the multi-component system. Although the detailed information of components is masked by aggregated states, lumpability is attainable with the partitioning rule. It means that the aggregated process is stochastically equivalent to the original one and retains the Markov property. We apply this model on a heat exchanging system in oil refinery company. The results show that fault propagation has a more significant impact on the systems lifetime comparing with inherent dependence and induced dependence.

Identifying the Requirements for Resilient Production Control Systems

Tighter supply chains create an increasing need for manufacturing organisations to become more flexible and more able to cope with disruptions drives the need for resilient production. Further, the interconnected nature of production environment and the complexities associated with the adoption of lean and process automation requires monitoring and control system to have increasing functionalities. Beyond simple monitoring and control, production systems are required to analyse information from disparate sources, detect abnormal deviations and then to react and cope with those deviations in a more effective manner. In this paper key requirements for resilient production systems are developed by establishing the links between production disruption and the required resilient capabilities. This then translates into requirements for resilient control and tracking in production systems.

Smart tracking to enable disturbance tolerant manufacturing through enhanced product intelligence

There is increasing need for manufacturing organisations to implement lean, just-in-time, make-to-order systems, mainly due to the cost pressures and varying customer preferences. This creates unexpected disturbances within the manufacturing systems, causing delays in delivery time. In order to quickly identify and react to disturbances, it is vital to capture real-time dynamic information related to the parts in production, resources, inventory levels and quality information. In many literatures, product intelligence achieves the fundamental requirements of managing disturbance. Current challenge, however, is that existing researches focused on developing a tracking system dealing with specific disturbance. In this paper, we present a systematic guideline for implementing such a system. The proposed guideline uses principles of product intelligence and combines them with the characteristics of disturbances and the associated information requirements. A case example is also presented to illustrate the developed concepts.

A Maturity Framework for Operational Resilience and Its Application to Production Control

This paper is concerned with resilience and its role in the operations of industrial processes. We refer here to operational resilience as the ability of an (industrial) operation to respond and recover in the face of unexpected or uncontrollable disruptions. The aims of this paper are to provide a common framework for examining the different challenges associated with assessing and improving operation resilience (b) Identify a set of levels for assessing operational resilience capabilities which can enable the positioning and comparison of initiatives taken to assess and improve it. (c) To illustrate the use of the operational resilience framework in the case of a laboratory forming and assembly operation.

Modelling food sourcing decisions under climate change: A data-driven approach

Abstract Changes in climate conditions are expected to pose significant challenges to the food industry, as it is very likely that they will affect the production of various crops. As a consequence, decisions associated with the sourcing of food items will need to be reconsidered in the years to come. In this paper, we investigate how environmental changes are likely to affect the suitability and risk of different regions—in terms of growing certain food items—and whether companies should adapt their sourcing decisions due to these changes. In particular, we propose a three-stage approach that guides food sourcing decisions by incorporating climate change data. The methodology utilises environmental data from several publicly available databases and models weather uncertainties to calculate the suitability and risk indices associated with growing a crop in a particular geographical area. The estimated suitability and risk parameters are used in a mean-variance analysis to calculate the optimal sourcing decision. Results from a case example indicate that sourcing decisions of popular food items are likely to require significant adaptations due to changes to the suitability of certain regions.

Customisation in Manufacturing: The Use of 3D Printing

An increasing demand to provide customised products creates challenges for manufacturing organisations. This poses a need to understand the characteristics required for manufacturing systems to handle customisation. In this study, 3D printing technology is assessed as an enabler for customisation. Additionally, the requirements of manufacturing systems with respect to configuration and control co-ordination are explored. A demonstrator is implemented to integrate 3D printing with conventional manufacturing, using an agent based distributed control system that co-ordinates the customisation of products and the order management.

Repair Services for Domestic Appliances

There has been a trend of increasing levels of Waste Electrical and Electronic Equipment over the last few decades as the possibility for accessing repair of appliances has declined. Reducing prices of appliances has also generated a culture where the disposal and replacement with new appliances is the quicker and cheaper option, compared with repair. A number of key areas have been identified as important in helping to increase the number of appliances which may feasibly be repaired in the future. Of these, two key areas encompass the automation of repair of appliances, and the information requirements in order to achieve this. Within this paper, a demonstrator will be described which provides a step towards illustrating the potential of product intelligence and semi-automated repair.

Towards Value-Based Asset Maintenance

The management of assets such as equipment and infrastructure can be a challenging task, and optimizing their usage is critical. Consequently, the importance of the maintenance function has increased because of its role in ensuring and improving asset performance and safety. Over the past few decades, there has been increasing interest in the area of maintenance modelling and optimization. More recently, the focus of research has been to take a whole-life perspective of the asset, and optimise maintenance decisions across the complete asset lifecycle. Most of the current research in this area takes cost (e.g., Life Cycle Cost) as the primary objective for optimisation (minimisation). These approaches do not effectively represent the role of maintenance because they do not consider the performance improvement organisations can expect to gain by proper maintenance – and this, we feel is a key limitation. In order to use maintenance as a “value driver” for the organisation, one must move away from cost-based thinking to value-based thinking. An important step in this direction is to consider net present value/utility of the decisions as the objective function, which will be discussed in detail in this paper. Nevertheless, the key parameters that are involved in a NPV or MVA based optimisation are still related to cost or “money” in general. By doing so, we would miss a number of other value-drivers that would be affected by maintenance. Examples of these are quality of products/service, customer satisfaction, environmental impact, etc. In this paper, we examine the possible elements of value provided by assets to the organisation owning those assets, and discuss how these value-drivers are affected by maintenance decisions. In this direction, we propose a measure - Value of Ownership (VOO) - to assess the value of maintenance and performance of maintenance decisions throughout an assets lifecycle. This measure will consider different value-drivers of a decision and makes it possible to consider the impact of maintenance decisions on a broader value space.