MICE: a facility maintenance expert system

This paper describes the implementation of a performance analysis (PA) review and a comprehensive root cause failure analysis (RCFA) for the artificial lift system (ALS) failures in heavy oil wells. The subject wells, located in Egyptian Eastern Desert (EED), have been operated by Canadian-Egyptian joint venture. Progressive cavity pump (PCP) and beam pump (BP) have been selected as the preferred ALS methods. With the objective of improving ALS performance, RCFA identifies the main causes of ALS failures and proposes guidelines and recommendations with a new open source service for the current and future development for EED Oil companies. PA has been applied using some Key Performance Indicators (KPI), such as: failure index, recurrence index, and average run life in order to track all failed and crucial issues. RCFA has been applied to different ALS, such as: BP and PCP. Therefore, the definition or limitation of the boundaries of each system and the classification of failures were performed. RCFA covered failures between 2012 and 2018 for average 150 active wells and 622 failures. Finally, the new open source service applied as a trial to evolve the failure tracking and decision making methodology. PA review showed a high failure and recurrence index that reached 3.0 and 3.7, respectively, in 2014 then declined to 1.0 and 1.7, respectively, in 2017 as a result of comprehensive corrective actions implemented. RCFA showed that rod string failure in BP was and still is a dominant failure with 200 failures over the last six years. Moreover, Down-hole pump failure in PCP was the major PCP failure system although its failures declined from 80 failures in 2014 to 14 failures in 2018. Several field cases were presented to cover the dominant cause of failures for both BP and PCP wells. Most wells had a significant improvement in their mean time to failure (MTTF) after reviewing design, installation and operation procedures. The procedure for carrying out this methodology and implementing lessons learned has been presented in this paper. In addition, the new open source integration service provides an increased visibility about individual well performance issues and more broadly, about field performance and ALS failures.

Empirical studies in software reliability research have predominantly focused on end-user applications. In this paper we present an empirical study on operating system reliability, which considered a sample of 7,007 real operating system failures collected from 566 computers. Our approach considered the clustering of different types of failures, as well as used an algorithm we created to search for and classify the causes of failures. We performed quantitative and qualitative analyses on the failure patterns and causes found, in order to investigate their correlations. The findings showed that failures in operating system services were dominant, especially related to software updates. We also found strong evidences of failure correlation, a property that is often neglected by studies in software reliability. In particular, we found evidences of causality relations between different types of operating system failures.

The development of a rule-based expert system (referred to as SCAR) for assisting electricians in diagnosing electrical faults in a shuttle car is discussed. Insight 2+, a microcomputer-based expert-system development tool by Level Five Research, Inc. was used in creating the system. The program requires the user to specify the initial symptoms of the failed machine; then the most probable cause of failure is traced through the knowledge base, with the software requesting additional information, such as voltage or resistance measurements, as needed. A causal-reasoning approach, which was used to develop the production rules, is described. The organization, implementation, and limitations of the knowledge base are also presented, along with a brief discussion of the ongoing development of a new system, SCAR.2. >

Determining the correct mechanism of a boiler tube failure is important for the prevention of future tube failures. Proper corrective measures can be undertaken to alleviate the root cause or causes for a failure only when the correct mechanism is known. A mechanism of boiler tube failure is defined as the process which degrades the tube and produces a failure. An expert system package was developed for use by electric utility power plant personnel during investigation into the cause of a boiler tube failure. It is intended to assist their efforts to recognize and identify a basic failure mechanism that can produce the type of boiler tube failure under investigation. The program guides the user to the most likely failure mechanism by asking simple questions, answers to which can be obtained by determining the failure location, type of fuel, and from visual inspection of the failued tube. Expert systems are a form of artificial intelligence which imitate the decision-making process of experts.

Despite decades of research in extensible operating system technology, extensions such as device drivers remain a significant cause of system failures. In Windows XP, for example, drivers account for 85% of recently reported failures.This article describes Nooks, a reliability subsystem that seeks to greatly enhance operating system (OS) reliability by isolating the OS from driver failures. The Nooks approach is practical: rather than guaranteeing complete fault tolerance through a new (and incompatible) OS or driver architecture, our goal is to prevent the vast majority of driver-caused crashes with little or no change to the existing driver and system code. Nooks isolates drivers within lightweight protection domains inside the kernel address space, where hardware and software prevent them from corrupting the kernel. Nooks also tracks a driver's use of kernel resources to facilitate automatic cleanup during recovery.To prove the viability of our approach, we implemented Nooks in the Linux operating system and used it to fault-isolate several device drivers. Our results show that Nooks offers a substantial increase in the reliability of operating systems, catching and quickly recovering from many faults that would otherwise crash the system. Under a wide range and number of fault conditions, we show that Nooks recovers automatically from 99% of the faults that otherwise cause Linux to crash.While Nooks was designed for drivers, our techniques generalize to other kernel extensions. We demonstrate this by isolating a kernel-mode file system and an in-kernel Internet service. Overall, because Nooks supports existing C-language extensions, runs on a commodity operating system and hardware, and enables automated recovery, it represents a substantial step beyond the specialized architectures and type-safe languages required by previous efforts directed at safe extensibility.

This paper will discuss some of the issues involved in building an Expert System that embodies tuning rules for IBM's MVS/XA operating system. To understand the components of an Expert System and their functions, PROLOG on an IBM PC (Turbo-PROLOG from Borland International) was chosen as the development environment. The paper will begin by defining the key concepts about Expert Systems, Knowledge Engineering, and Knowledge Acquisition. The reader will be given a brief overview of PROLOG, from which we can explain how an inference mechanism was developed. Finally, the paper will describe the Expert System that was developed, and additionally will provide a set of key issues that should be addressed in the future. It is our overall objective to provide new insight into the application of AI to CPE.

This paper studies the reliability measures of a system consisting of two subsystems in a series configuration for different types of failure and two types of repair. The subsystem-1 has four identical units in a parallel configuration operating under 3-out-of-4: G policy and this has connected to subsystem-2 which has three identical units arranged in a parallel configuration and working under 2-out- of-3: F, scheme. The units of subsystem-1 are controlled by a controller for preventing failure effect and safety purposes. It is assumed that units of each subsystem have different types of failure and repair rates. The unsuitability of the environmental conditions such as overheating as a well-known natural cause of failure of any system and also weather conditions like heavy rain, thunderstorm, and catastrophic shakeups, etc. have treated as a complete failure of the system. This study considers the environmental causes of failure in the proposed repairable system as complete failure by which the system stops functioning. Human failure in the system is trickled as complete failure and repair employing copula (Gumbel- Hougaard family copula distribution) like another complete failed states of subsystems. To analyze the proposed system, the supplementary variable technique are used and some measures of system reliability like availability, reliability; MTTF and incurred profit function for different values of parameters are derived. Some particular cases for different values of failure rates that have explicit are also studied.

To improve the performance of fault diagnosis expert system based on ANN IN fields of convergence speed, locally optimal solution and the low accuracy, an missile fault diagnosis expert system based on GA-BPNN is proposed in this paper. The genetic algorithm (GA) is adopt to optimize the weight and threshold of matrix while BP neural network realizes the non-linear map relations between failure feature and failure cause. The simulation results indicate that the method proposed in this paper significantly increase the convergence speed and globally optimal solution of neural network, the fault diagnosis accuracy of expert system for a missile has been improved also.

As current backbone network evolution involves replacing today's multiple networks with a single global multiprotocol label switching (MPLS)-enabled backbone over an intelligent optical IP-based core network, alarm correlation becomes critical for fault management system (FMS) to monitor network health, performance, and to quickly identify and resolve operational problems. In this paper, we propose a hybrid technique to efficiently correlate these MPLS alarms to other equipment and service alarms, including event aggregation, rule-based method, and codebook approach. The hybrid technique combines the advantages of small size codebooks and simple expert rules for correlation at each level of the hierarchy. Extensive testing results obtained from real-world network experiments have been reported to demonstrated the effectiveness of the proposed correlation scheme.

In order to improve the operational efficiency of decision-making in the context of spacecraft (SC) endurance in operation, the task was set to increase the efficiency of adaptation of its control system to the environmental effects. The instruments installed on most Russian SCs in many cases do not provide for identification and timely elimination of accident sources due to delays in the identification of faults and failures. A technology is proposed that involves intellectualization of control systems. It is suggested to complement the SC control circuit with an expert system that includes a “prognostic decision support system”, a “control simulation and correction module”. Due to the ambiguity and common uncertainty of cosmic phenomena, it is suggested to predict the reaction of SC equipment to external effects rather than monitor such effects. The intelligence of the expert system is to be ensured through the analysis of the communication medium that defines the possibility to ensure SC survivability. The correction of control is suggested to be performed not on the basis of process parameters monitoring, but rather knowledge. This knowledge is held by experts who possess experience in SC flight mission performance. The results of the audit of external factors and development of SC functional units reactions represent the input data. After clearing, sorting and statistical analysis of data, it is suggested to regard it as information resources. The results of such resources analysis and design of messages based on expert conclusions transforms such resources into knowledge that is used in decision making and control correction. The diversity of architectures and processes of SC functional units design has defined the requirement to involve experts with diverse professional backgrounds. It was proposed to generate forecasts of SC equipment reactions development in the form of description of the dynamics of multifactor combination of the results of intersubject audit of functional units operations and subjective expert evaluations. In order to ensure agility of information analysis within the knowledge base, it was suggested to use the OLAP comprehensive multidimensional analysis technology. In particular, that regards fast analysis of shared multidimensional information that includes requirements for multidimensional analysis applications. The proposed model of systematic accumulation and processing of knowledge will enable flight control officers to timely identify inadequacies in the control inputs. The logical and statistical analysis capabilities ensured by this application will enable the delivery of analysis results to the experts within a time period sufficient for elimination of the causes of faults and failures in SC equipment operation. Multidimensional conceptual representation of data including the support of multiple hierarchies will define the capability to refer to any required information regardless of its size and place of storage. The proposed method of information support of SC equipment reactions forecasting is addressed in the light of the analysis of electrophysical effects that affect SC in near-Earth orbits. Combining the methods of computer data processing and intersubject analysis of functional units operation must insure efficient decision-making based on increased accuracy and agility of data processing and, consequently, selection of SC operation adaptation scenario subject to flight control officer’s preferences.

The metallurgy and levels of protective elements in the new P91 and P92 martensitic alloys are quite different to older ferrite pearlite and ferrite bainite steels. This is likely to give problems in the positive identification of failure mechanisms and assessment of the life of superheater tubing. Techniques such as spheroidisation and oxide thickness to assess metal temperatures, or hardness testing and cavitation to estimate the level of creep damage cannot be used with these materials, or are likely to give ambiguous results. These issues are discussed in the paper and were explored in depth in part of a project at the University of Surrey to develop “If-Then” rules for an expert system for power plant failure analysis.The main aim of the proposed expert system is to identify the root causes of such failures in superheater tubing and will encapsulate current knowledge about superheater problems in the form of “If-Then” rules. Root causes of creep failures include furnace design and operation, overestimation of alloy creep properties, inadequate heat treatment and a non-optimum content of strengthening elements.A characteristic of these alloys is that oxidation on the steam side of the tubing can induce premature failures due to the insulating effect of the oxide scales raising tube temperatures. In addition, scale spallation could also increase tube temperatures, as spallation debris may collect in the bottom of tubes, blocking steam flow. Attention is drawn to a potential “runaway affect” in which the tube temperature and rate of oxidation increase with time as the oxide builds up. The root cause of this could either be excessive rates of heat transfer or could be due to inadequate oxidation resistance caused by low levels of protective elements. Methods of identifying which of these is the root cause are discussed.

The control centres of New Zealand's power transmission network arc flooded with alarm messages during a major system outage. A majority of these faults involves the transformer trips. To help the system controllers to restore the transformers Transpower NZ Ltd., New Zealand's power transmission company, has developed a expert control system to perform a real-time analysis of the incoming alarms, identify transformer related events, analyse the faults and provide a real-time restoration advise on transformer restoration. This paper describes the problem tackled by Transpower's system controllers and our expert control system that assists in solving the problem. It discusses the types of knowledge represented, the architecture involved, some implementation details and the performance analysis.

The importance of operators' knowledge in the production of high-quality water and the need for consistent high-level expertise have fueled propositions for the application of computer technology, including knowledge-based expert systems (KBES), as a means of enhancing drinking water treatment operations. With this in mind, the authors have developed CHLOROS, a knowledge-based system which integrates both expert system and hypertext technologies in a WindowsTM environment. Designed as an off-line tool, the system does not directly control operations related to drinking water disinfection, but rather has been intended to assist treatment plant operations personnel. It is believed that CHLOROS will serve as a practical tool for water utilities because it creates a consistent decision-making environment, captures and preserves the expertise of human experts, and reduces the amount of manual effort and time that may be expended by searching through manuals and literature for information that is required to support the choice of control parameters.Key words: knowledge-based systems, expert system, hypertext system, disinfection, chlorination, chloramination.

In this paper, the research area of real-time monitoring and fault diagnosis of locomotives is introduced. Also included are the collecting, categorizing and analyzing the real-time data of the on-board equipment, as well as the application of new technologies such as the real-time data transferring between the on-board equipment and off-board equipment, the fault warning and the analysis of the fault changing trend, and the automatic generation solutions for the fault. The research of the collection, categorization and analysis of the locomotives and the onboard equipment is studied deeply. Combined with the situation of the Chinese railway, a database of locomotive fault is constructed. And by the expert system and information technology, codes and the description of the real-time fault information can be fully collected. Based on a preliminary assessment of fuzzy matching on the fault library, a solutions database is constructed. Finally, the benefits of the expert system in safety operations and locomotive maintenance are summarized.

In view of the complexity of the electronic control system of adaptive optics, combining with fault tree analysis and expert system, we design an expert system on fault diagnosis for electronic control system of adaptive optics based on Fault Tree. After analysis of electric control system of AO, fault tree can be established. The expert system is the extensive application of the fault tree. The system uses forward inference method to diagnose failure of system. In addition to traditional expert systems, point of failure was set to obtain real-time fault information. Then the information was sent to diagnosis expert systems to get the online function of diagnosis of system. The system highly improved the stability of AO system.