4.2. Decision-making models

Strategic, tactical, and operative decisions are made on the various aspects of business operations. The vision of an industrial enterprise must take into consideration the changes in its operational environment and the leadership practices. Decision-making is supported by analyses, models, and computer-aided tools [30,31,32]. Technological advances have an impact on the business of industrial enterprises and their use of new innovations. Industrial innovations contribute to increased productivity and the diversification of production and products, they help to create better, more challenging jobs and to minimize risks. This study aims to improve decision-making in questions of field technology towards better productivity and efficiency [75,76,77,78]. Making good decisions can be enhanced by goal-oriented, analytical activity of the users of field technology aimed at maximising the use of technology. On the other hand, decision-making is complicated by various factors, including external aspects: the decision-maker is faced with a number of possible choices, and must fit together the activities of the enterprise and environmental considerations [17,79,80].

In this section, we take a closer look at the decision-making process. Its classifications will first be discussed, followed by an analysis of how decisions are made, and finally we will take a look at various models for decision-making. Siddall has classified the process from a problem-oriented perspective, using the following categories: the choice of design option, design optimization, specification, productivity factors, and risks. All of these problems are also encountered when making decisions concerning field technology [81].

Long-term decisions have an impact on process changes, functional procedures and maintenance, and also on safety, performance, costs, human factors and organisations. Short-term decisions deal with daily actions and their risks. Decision-making is facilitated by an analysis that incorporates a classification of one’s own views, calculating numerocal values, translating the analysis results into concrete properties, and a numerical evaluation of the properties. One method applied for this purpose is the Analytic Hierarchy Process (AHP) model. This model, which has many features in common with the multivariable decision-making model (MCDM) applied in this research, is also suited for field technology decision-making that aims at making the correct choices both in the short and in the long term (cf. Chapter 7.1) [82].

Holmberg et al. [77,78,83] describe decision-making and decision analysis in the same way as Fig. 10. In order to combine analysis and technical knowledge, an analysis of consequences and usage is used together with a probability estimation of safety factors. Decision analysis has the benefit that it gives information of facts and alternatives, in addition to calculated estimates. Quantitative and qualitative factors, the extent of knowledge, and the analyzed information are evaluated by means of various approximation methods, and the safety of the options is estimated using probability calculations according to the Probability Safety Assessment (PSA). The phases of decision analysis are [77,78,83,84]:

1. Structuring the decision problem and identifying decision alternatives

2. Defining the objectives of the decision

3. Defining performance measures or variables for the quantification of decision objectives

4. Identifying critical uncertain variables

5. Probability assessment

6. Specifying value judgement, preferences and trade-offs

7. Evaluating alternative actions or policies

8. Conducting sensitivity and value of information analyses.

Multivariable decision-making model. One example of multivariable models is the decision-making board. Its vertical columns list the decision-making alternatives, the horizontal rows the criteria or functional variables. Another possible model is a tree-shaped, multivariable decision-making model suitable for solving difficult problems. The advantages of this model include minimized costs and risks and maximized benefits.

The decision tree illustrates the changes and decision nodes, and it includes a limited number of variables. This procedure structure is applicable to both continuous and discrete cases. Figure 11 illustrates the tree-diagram model for deciding on the construction of a large factory [85]. Performance vectors (u_n) and probabilities (p_n) are used as variables in the calculation, and the model yields numerical benefit values (c_n). Risks are estimated at the change and decision nodes (r_n). The method presented by Lindley [85] contains exact calculations. In the present study no calculations are made, but calculatory application of the method could well be used when intelligent field technology is being diffused to factories [77,78,83,86,87,88].

Elomaa describes the use of the decision-making tree in learning. It enhances the motivation of teachers and facilitates the study of environmental opportunities [86]. The so-called ”machine learning” is used as a tool, applying practical tools and techniques [82]. One of the goals of the present research is to promote the diffusion of field technology. The multivariable model also promotes learning among decision-makers and end users in factories and is a useful tool in making decisions concerning field technology (cf. Chapter 9).

Multiple-criteria decision-making (MCDM). According to Agrell, MCDM offers the methodology for decision-making analysis when dealing with multiple objectives [89]. This may be the case when the success of the application is dependent upon the properties of the system, the decision-maker, and the problem. Problems with engineering design involve multiple criteria: the transformation of resources into artifacts, a desire to maximize performance, and the need to comply with specifications.

The methodology can be used to increase performance and decrease manufacturing costs and delays of enterprises. The decision-maker is an engineer able to use tools such as Computer Aided Engineering (CAE) and Computer Aided Design (CAD). Two important areas suited for modelling are structural engineering and electrical engineering. The Multiple-Criteria Decision Support System (MCDSS) uses the methodology and ensures mathematical efficiency. The system employs graphical presentations and can be integrate with other design tools. Modelling and analysing complex systems always involve an array of computational and conceptual difficulties, whereas a traditional modelling approach is based primarily on simulation and concepts taken from control theory.

The strength of the MCDM lies in the systematic and quantitative framework it offers to support decision-making. Comprehensive tuning or parametric design of a complex system requires elaboration on utilizing the modelling facilities of system dynamics and the interactive decision-making support of the MCDM. The multiple-criteria decision-making model applied in this thesis to support decision-making in the field technology of process management will be presented later (cf. Chapter 7.1) [87,90,91].