Experimental research, areas for research and reasons for developing expert systems

3. Experimental research, areas for research and reasons for developing expert systems

The basic facts of this research are attempted to define two levels of experimental data. The first level of the data is related to quality management systems and nonconformities that have emerged. This is a basic level of data which reflects the situation in the quality management systems and identify critical places that are subject to improvement. The base of these data is unique and consists of the 1009 nonconformities (cases), identified in over than 350 organizations. If we know that in our area in the field of competent certification body has, approximately 500 certificates, then the number of 350 is about 70% of the total number. That fact points out to the significance of sample for analysis.

The term nonconformities refer to any non-conformance of requirements of ISO 9001, nonconformity non-fulfilment of a requirement [21]. During the external audits of quality management system, competent and trained auditors can identify several types of nonconformities (Figure 1). We are using most significant data from highest level of pyramid at which were collected at the level of many country like external estimation and evaluation of they performance and condition.

Distribution of nonconformities depends on the rules that define the certification body itself. However, for the purposes of this research is used classification which is the most common in the literature, which is favour by the authoritative schools in the world in the field of management system and that is clearly recommended by European guidelines in the subject area, which is split into three levels. The first level is the disagreements that are evaluated as insignificant deviations from the standards and requirements which are interpreted as an oversight or random error. The other two categories are interpreted as nonconformities that represent a great deviation from the essential requirements, which are reflected in the frequent discrepancies in individual requirements, representing a deviation that brings into doubt the stability of the management system and threatening the operations of the organization.

Data base of nonconformities which is under consideration in this research contains only nonconformities in the domain of the other two categories, and that giving greater importance to this research and gives greater significance results.

Non-conformances are identified in accordance with the structure requirements defined in the ISO 9001 standard as follows:

- Quality management systems: 4.1 general requirements, 4.2 documentation requirements,

- Management responsibility (module 5): 5.1 management commitment, 5.2 customer focus, 5.3 quality policy, 5.4 planning, 5.5 responsibility, authority and communication, 5.6 management review,

- Resource management (module 6): 6.1 provision of resource, 6.2 human resources, 6.3 infrastructure, 6.4 work environment,

- Product realization (module 7): 7.1 Planning of product realization, 7.2 customer related processes, 7.3 design and development, 7.4 purchasing, 7.5 production and service provision, 7.6 control of monitoring and measuring devices,

- Measurement, analysis and improvement (module 8): 8.1 general, 8.2.1 customer satisfaction, 8.2.2 internal audit, 8.2.3 monitoring and measurement of processes, 8.2.4 monitoring and measurements of product, 8.3 control of nonconforming product, 8.4 analysis of data, 8.5 improvement.

Accordingly, for example in the field of 8.2.1 from the standpoint of the appearance of nonconformances organizations have a significant and frequent or large deviations in the sense that it does not follow the information about the observations of users, it did not define the methods for obtaining this information, they do not have strong communication with customers and similar. Or for example in the field of 8.2.3 with the observed aspect, organizations do not apply appropriate methods for monitoring and performance measurement processes, have not mechanisms for implementation of corrective measures in cases that have not achieved the planned performance of processes and the like.

This data will be used like the basis of CBR approach or approach where it is possible to make significant conclusion in the sense of main target of this research. This approach is shown in figure 2.

The second level of data consist data from evaluation organizations that participated in the competition for the quality award based on European Quality Award criteria. This database is unique, as well as in the previous case. Data were transferred in encoded form in order to secure the identity of the organization. Data were collected in 100% extent (34 organizations) and thus are significant and give a real picture of the situation in our organizations. These data are used for comparison with previous, basic level data. That is way for making improvement or exalt from basic level on the level of business excellence and way for making knowledge which reproduce expert system on his output. That is also comply with literature more existent attitude, and natural way that organization should first implement Management concept [22, 23-26, 27].

In order to show the current directions and trends in the field of development of software for quality, and to select under researched areas in the field of software quality, it was conducted a detailed review and analysis of a total of 143 software. All necessities information for that analysis are available in site (http://www.qualitymag.com) where are publish updated software items which are related to quality. The results of the analysis are shown in the figure 3.

On the x axis diagrams are shown the software ability and orientation. Obviously is that the software in the field of quality is usually oriented to the control of documentation, statistical control and analysis, six sigma model, concept of total quality management, FMEA and QFD methodology, corrective action, flowchart and process mapping. However, there are specific tools for automation: the implementation of the quality management system documentation, description of information flow, implementation methods and techniques of quality, and more. Therefore, it can be concluded that there is no software that is based on the application of artificial intelligence tools in the sense of the definition of preventive actions for the purpose of improving the process. The greatest number of software is related to the application of statistical methods in the process of monitoring and improving quality. It is obviously that a large number of software is based on total quality management systems concept. The facts point out present approach which we develop in this research

and also justify further research in this area. It is interesting that a large number of software are base on the corrective actions and on the other hand there is not any registered software that has application for output preventive action what is, of course, main recommendation of ISO 9000 series. This fact also gives stimulus in terms of development of software that emphasis to the prevention. That approach is unique in the field of software for quality and makes this research more significant.

Beside this analysis, in this research were analyzed huge amounts of available books in order to point out the justification of applying expert system. Expert systems are different from other artificial intelligence systems in that, they attempt to explicitly and unequivocally embody expertise and knowledge with the software [28]. Expert systems are also identified as one of the most commercial branches and in most number of projects used artificial intelligence tools [29, 30]. For example, it is estimated that in the first half of 21st century, even 75% of all legal documents be written with the assistance of expert systems [31]. Also expert systems will be of vital importance for measuring the quality of products and services [32-34]. Expert systems are an area of special importance with rise trends in modern business conditions [35, 36-38]. They have special significance in a highly developed countries where is actual knowledge based economy. This research highlight trends,

significance and justification of developing and implementing expert systems.

Main idea and approach for developing expert system come from analogy between human body functions and process in some organization which was organized based on process modelling from ISO 9000 respect. This approach is present on figure 4.

This research tries to deal with perfection of functioning of the human body compare with a process modelling structures of the implemented quality management system. The challenge made in this way, tried to create a system that is universal for all sizes of organization, which incorporates a large number of gathered data, in fact a large number of experiences, in order to get a better image of the system status. This should be added to the primary goal which is to develop a model for improvement of management system, oriented to achieve BE according to show off how to maintain and improve the performance of the human body. However, the goal is also, to develop a system for measuring

performance and capacity of each activity in the QMS, in order to obtain a true picture of the systems and capabilities in order to define the areas where improvements should be made, with clearly defined intensity of improvement. On the basis, thus established the analogy is made to compare elements of implemented QMS to the systems that have applied for Quality award for BE as a system with high performance.

To establish the analogy between the process modulated organizational structure and the human organism, so as to create the system that is independent from organizational functions and based only on the process model, following division of man functions was made [39, 40]:

- willing and

- unwilling functions.

Willing functions (term “functions” is used in medical terminology, although it is equally correct, to use a term “activities” in view of ISO 9000 standard terminology. For reasons of consistent referencing and use of theories from the field of medicine, the author has chosen to use the term functions.) are those dependent on man’s profession and performed by man’s will. They are variable and dictated by a central control of the organism. For example, when a worker at the construction site lifts his hand, it is not the same as when a referee at the game lifts his hand and etc. Willing functions refer to functions of external motoric organs.

Second category is made of unwilling or automated functions and their use is given by their existence. There are functions that are same in all professions and all people (considering that they exist, i.e. that human body is in good health) and do not depend on the man will but are simply executed. For example, those are functions of secreting enzymes, hormones, heartbeats, and similar, like ordinary body functions, and functions that cannot be controlled [41, 42].

With such a ratio of functions in the human body, we can establish the analogy of the system with implemented quality management system. Analogy in term of willing function goes in direction to developed all data in to two category, production and service organization and make some analyses, which is not subject of this research.

In order to meet requirements of this research, only analogy in terms of unwilling functions has been considered. The idea is to use all nonconformities (undependable of organization type or size) and base on case base reasoning approach, make conclusion about readiness of systems to making some top form.