Analysis and development of technology's value model for machine tool renewal management.

Albu, Adriana ; Ruset, Vasile ; Stanciu, Loredana 等

1. INTRODUCTION

In Europe, manufacturing is the dominant sector--with the signification of more than 40% in value added jobs. In the Manufuture strategy, two guidelines appear (Jovane & Carpanzano, 2007):

* compete by reducing costs (cheap labour, automation);

* compete by added value (high performances, customization, new business models, new human capital, and service dimension).

For a manufacturing company from mechanical industry, one of the key factors is the machine tools performance (Childs et al., 2000). The problem statement is what are the practical guiding principles for the management to acquire or renew a machine tool. Success of a firm can be very good evaluated by the superiority of the machine tools. However, this parameter alone can conduct to incorrect conclusion, if is not correlated with the manufacturing environment. Considering the machine tool as a complex product, the specific and simplified Life Cycle Management algorithm is shown in figure 1.

[FIGURE 1 OMITTED]

The closed loop manufacturing: renewing functions while circulating material (Tani, 1999; Dornfeld, 2007) is in reality further complex, although for our purposes we can adopt the model from figure 1. First, we define the technology's value of a machine tool as shown in equation 1:

[nabla] [MTT.sub.T] =f ([P.sub.T] + [C.sub.T] + [F.sub.T]) (1)

where [P.sub.T]--technologically production capability for a given part at the moment T; [C.sub.T]--available capacity for a given part at the moment T; and [F.sub.T]--financial efficiency of a machine tool for a given part at the moment T.

Naturally, the given part can have a variety of complexity. The measure of a part complexity CX is related with the number of its independent dimensions and the precision to which they must be made as is shown in the equation 2 (Childs et al., 2000):

[C.sub.x] = n x [log.sub.2] (1/[nabla]1) (2)

where n is the number of the dimensions of the manufactured part, and (l/[nabla]l) is the average fractional precision which have to be produced.

2. THE TECHNOLOGY'S VALUE MODEL FOR A MACHINE TOOL

The machine tool technology's value is variable in time, as is shown in figure 2. Machine tools are capable of higher productivity, but they are also more expensive--the correlation is given by different authors (Krar et al., 2004). To simplify the algorithm, in this paper we consider one machine tool manufacturer and one machine tool type (i.e. a milling machine). In time, the manufacturer continuously develops the product (the milling machine). The machine tool technology developed by the manufacturer is shown on the segments 0 - 1 - 2 - 3 -4 and so on. A company will acquire the milling at the moment [T.sub.1].

[FIGURE 2 OMITTED]

From this moment, the [MTT.sub.1] value of the machine tool will decrease continuously, affected by the depreciation D. The company possibly will renew the milling machine at the moments [T.sub.2], [T.sub.3] or later. After the financial analysis, considering the technical improvements shown in figure 2, the management will assume the appropriate decision.

3. TECHNICAL SOLUTION: AN ARTIFICIAL INTELLIGENCE MODEL

The manufacturing process and renewal decision based on technology's evolution model is exposed in figure 3. Financial analysis depends on balance sheet and income statement, resulting analysis ratios--profitability and activity (Bruzzone et al., 2007).

[FIGURE 3 OMITTED]

For the machine tool renewal management the research course implying artificial intelligence has three possible decisions (suggestions):

* [S.sub.1]--maintain the same machine tool technology, and delay the minor improvements according with actual technology;

* [S.sub.2]--expand the MTT value for the existing machine tool applying the necessary updates;

* [S.sub.3]--develop an ultimate technology value by acquiring a new machine tool.

An expert system based on logical inference was developed in order to make a suggestion regarding MTT value. This type of systems is used to implement human reasoning (a virtual company's management) and for doing this, it needs a set of premises (properties of the parameters that describe the machine evolution from different points of view: technologically production capability, available capacity, financial efficiency). The premises will form, together with logical operators, the rules of the system. These rules create a knowledge base that is the most important element in logical inference. According to the rules defined in the knowledge base, the elements that characterize MTT value are interpreted and the decision is made.

The data structure used in logical inference is an oriented graph. A chain logical evaluation, given by the rules, is applied on this structure.

In this case, the premises are strongly connected to the machine tool technology at a moment T. The parameters, which describe [MTT.sub.T], are analyzed in the equation 1.

The system compares, from technical point of view, the parameters of the actual machine with those that characterize the decision Sx. After that, the financial analysis is made. All these actions are involved in decision-making and using them, the rules of the expert system are made.

The general rule's form is: "If logical_premises then conclusion", where logical_premises are explained by the equation 3.

[V.sub.2] - ([V.sub.1] - D) > [Q.sub.UP] (3)

where [Q.sub.UP] is the value of the machine tool upgrade investment [S.sub.2] or [S.sub.3]. The knowledge base (in fact the rules of the expert systems) is machine tool manufacturer dependent. Besides, technology development evolution for each manufacturer in the knowledge base must be predicted by our artificial intelligence system. The most difficult part of implementing expert system was to obtain the prediction curves regarding the technological evolution for the machine tools manufacturer (see the segments 0-1-2-3-4 from figure 2).

In the construction of the graph structure, which represents the knowledge base and the support for the decision making system, two groups of information were used: (i) the manufacturer source as primary information, and (ii) additional scientific researches (conferences, seminars, symposiums, etc.). Also, essential information was collected from hardware and software industries connected to machine tools domain (i.e. Sinumerik[R], Fanuc[R] controllers).

For the first version of this artificial intelligence expert system InterProlog (a Java front-end and enhancement for Prolog) was used to develop the knowledge base. It was chosen because of its mutual recursion and multithreading support. These basic requirements were absolute necessary because of the great number of parameters simultaneously computed. For the milling machine case 32 parameters were considered.

4. CONCLUSION

The new results of this paper are: first, we created a model for machine tool technology, in order to evaluate technical and financial level. The decision to upgrade the machine tool or to renewal it completely is very hazardous for a manufacturing company--the user of a machine tool. Several parameters are involved in this decision, and decision usually is taken by the management, assuming an extreme risk. An artificial intelligence system was billed as a software application to support the complexity of the decision. As further works, the AI model must be refined and extended to other machine tools--our first model was created only for CNC milling machines.

5. REFERENCES

Bruzzone, A.; Lonardo, P. & Rossi, E. (2007). Production Process and Inventory Management for Machine Tool Companies: Analysis and Development of Optimisation Models, International conference on smart machining systems, NIST, Gaithersburg, March, 13-15, 2007, Available from: http://smartmachiningsystems.com/ slides/nistBruzzone1D.pdf, Accessed on: 2009-02-10

Childs, T.; Maekawa, K.; Obikawa, T. & Yamane, Y. (2000). Metal Machining. Theory and Applications, John Wiley & Sons Inc., ISBN 0 470 39245 2, New York

Dornfeld, D. (2007). Sustainable Design and Manufacturing: Can we "Engineer our way" to a Sustainable Future?, Available from: http://www.citris-uc.org/files/ energy_research, Accessed on: 2009-05-22

Jovane, F. & Carpanzano, E. (2007). Research and Innovation in Manufacturing Systems in Lombardia: from regional to inter-regional and European successful activities, Biregional Workshop Lombardia-Baden-Wuerttemberg, Stuttgart, January the 29th, 2007, Available from: http://www.regstrat.net/download/stuttgart/ 2007-01-29_jovane.pdf, Accessed on: 2009-04-16

Krar, S.F.; Gill, A.R.; Smid, P. & Krar, S. (2004). Technology Of Machine Tools, McGraw-Hill Higher Education; 6th edition, ISBN 0078307228

Tani, T. (1999). Product Development and Recycle System for Closed Substance Cycle Society, First International Symposium on Environmentally Conscious Design and Inverse Manufacturing, pp 294-299, ISBN: 0-7695-0007-2, Tokyo, Japan