Expert systems in industrial practice: advantages and drawbacks

An important aspect of the production of quality software is that of software assessment and certification. This involves the measurement of a finished software product using a set of recognised Software Assessment Units. It is essential that at the specification of a piece of software a set of relevant software attributes with a suitable subset of Assessment Units is identified. The set is used to determine whether or not the finished product meets an appropriate quality standard. It has been recognised that a Software Assessment Advisor (SAA) is a necessary tool to assist with this task. An SAA can help in determining the software attributes which should be measured and in selecting an appropriate set of Assessment Units. The SAA will use information concerning the categorisation and functions of a proposed software product, the user's requirements, available tools, and prior assessment experiences in these tasks. So far, no clear approach has been identified for the development of an SAA. The investigation of the modelling required by an SAA indicates that Expert Systems techniques could provide a useful step forward in this development. Expert Systems techniques are able to relate the information required by the certification procedure intelligently by means of the knowledge model and to support symbolic reasoning to advise the software certifier. Also an Expert System's capability for knowledge learning is significant with respect to the relatively young area of Software Assessment and Certification. This paper discusses the activities involved in the Assessment and Certification process, describes how an SAA can assist for the process and how an Expert Systems approach has been used in the development of a prototype SAA.

Knowledge-based operation and management of communications systems

8 - Real-time fault-tolerant control systems

Decreasing supporting resources and increasing complexity of horticultural crop production necessitate the new technologies be developed to transfer information to farmers as well as commercial producers. Expert systems, a sub area of artificial intelligence of computer science offer the possibility to implement horticultural knowledge into field practice in the form of computer programme in order to support the farmers in decision making at farm level. This paper provides an overview of the concept of expert system and examples of areas of horticultural applications of these systems are in use at present are also outlined. This paper further examines how expert system techniques are being used at present in problem solving and why someone in the field of horticulture might want to use them.

Diagnosing the vibrating behaviour of the PWR internal structures using expert systems techniques and algorithmic treatment

As environmental regulations become more stringent, manufacturers are confronted with the of productdesign considering environmental impacts. Life Cycle Analysis (LCA) is a developing technique that attacks this problem. LCAs provide environmental information for decision making by consumers, manufacturers, and governments. These decisions significantly affect the ability to maximize source reduction, re-use, recyclability, and recycled content. LCAs can aid in diminishing the negative impact a product has on the environment and can be utilized as a partial solution of our nations' (worlds') growing waste management problem. To date, the application of LCAs has varied from product comparison to process development and improvement. The results from these analyses have differed because of differing assumptions and differing choices of boundary conditions. An appropriate Life Cycle Analysis should be consistent across products, locations, and especially across experimenters. The development of LCAs lends itself to computer-assisted design methods such as the techniques of Expert Systems, due to the large amount of information associated with successfully developing an LCA. In this paper, we will discuss using Hierarchical Classification, a specific task specific architecture expert systems approach, to organize the terms necessary in performing an LCA. Although not yet definitive, we believe our experience indicates that such approaches to organizing the vast knowledge necessary to accomplish an LCA may be a key factor in eventual successful development of comprehensive LCAs in industry.

There is a long-standing challenge to narrow the gap between software engineering research and industry practice, to align their interests and realize true synergies between the two communities. Some difficulties to this challenge include mismatched agendas, priorities and expectations from the research collaboration on both sides. To overcome these difficulties, an initial step is to gain a clearer understanding of collaboration challenges from both perspectives. With this goal in mind, we organized the 5th International Workshop on Software Engineering Research and Industrial Practice, collocated with the International Conference on Software Engineering 2018. The workshop featured two keynote talks, one from industry and one from academia, followed by paper presentations and a round-table discussion session. Here we summarize experiences shared by the keynotes from industry and academia, along with findings from paper presentations and overall discussions by workshop participants on the ways of aligning software engineering research and industry practice.

Contributors

Intelligent process-control system design

The aim of this paper is to present a Fuzzy Expert System Methodology (FESM), based on the Fuzzy Logic, Expert System techniques and multicriteria decision making theory, which can be used to calculate a unique Fuzzy Total Power Quality Index (FTPQI) for every measured point of the electric network (measurements for a specific time period are used). FTPQI of the specific point includes unified and concentrated information about its PQ level and could contribute to the evaluation of examined electric network nodes providing a ranking between them in terms of PQ using the corresponding FTPQI. This tool could be used by the system administrator giving a reliable and scientific way to check the total PQ of electric networks in general and at specified points and, also, providing the possibility of ad-hoc network improvement remedies. (6 pages)

Artificial intelligence and expert systems techniques, coupled with historical signal analysis and optimization methods, provide a novel approach to monitoring process dynamics and performance and learning knowledge about how to do a production job better based on past experience. An expert system-based supervisor is proposed to provide such a tool for a simple tank level control problem which adjusts the schedule for supply purchase to attempt to minimize the average cost of operation over time. This research provides a prototype for the applications of expert systems in real-time process supervision and control; however, the result is primitive. The underlying modeling and optimization methods only work for simple situations.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

The authors demonstrate how a pattern-recognition system can be applied to the interpretation of capacitance-voltage (C-V) curves on an MOS test structure. By intelligently sequencing additional measurements it is possible to accurately extract the maximum amount of information available from C-V and conductance-voltage (G-V) measurements. The expert system described, (CV-EXPERT), is completely integrated with the measurement, instrumentation, and control software and is thus able to call up a sequence of individually tailored tests for the MOS test structure under investigation. The prototype system is able to correctly identify a number of process faults, including a leaky oxide, as shown. Improvements that could be gained from developing rules to coordinate G-V, capacitance-time, and doping profile measurements simply by recognizing the important factors in the initial C-V measurement are illustrated. >

There are many environmental applications which rely on the accurate identification of insects, plants, etc. The availability of personnel with sufficient identification and taxonomic skill is often a limitation and the production of identification keys has been one means of alleviating this problem. It is the purpose of this paper to show that expert system techniques can be employed to provide computer based keys that are both simpler to use and easier to construct and maintain. This paper describes the structure, development and use of such an expert system based identification key. Experience indicates that quicker and more accurate identifications are possible and that there are other less obvious benefits. In particular the system can be both easily modified and maintained and that the addition of new knowledge is a relatively simple operation.

The expert system design of transmission line geometry is presented. The design focuses on the reduction of crosstalk and signal delay in microstrip lines. The first level of design utilizes the whole range for each design parameter and gives satisfactory cross section geometry as a result. The second level optimizes the geometry with respect to the chosen parameters. >

The authors propose a tool for systems analysis that incorporates hypertext, simulation and expert system techniques. The tool, Hyper Analysis Toolkit (HAT), provides a hypertext linkage of graphical models, such as DFDs (data flow diagrams) and ERDs (entity relation diagrams) with system description narratives and other documents created during the early stages of systems analysis. Hyperlinks placed in diagrams and documents provide an easy way for users and system analysts to navigate and cross-reference systems models. The tool also includes a simulation package and a rule-based expert system to evaluate models statistically and dynamically. Extensive evaluations of models will give estimates of the system dynamics at the early stages of the systems development life cycle.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>