On designing of layout of flexible manufacturing systems.

Ficko, M. ; Brezocnik, M. ; Balic, J. 等

Abstract: By proper layout of devices in the flexible manufacturing system (FMS) it is possible to reduce the variable transport costs and consequently, the total manufacturing costs and thus secure the competitive advantage. The paper describes the methods of placing of machines and devices in the FMS. So far, different methods of layout have been introduced, but the problem has not been satisfactorily solved yet.

Key words: flexible manufacturing system, layout, designing, optimization

1. INTRODUCTION

An FMS usually consists of several machines and material handling device. One of the key factors of the successful operation of the FMS is the transport between the individual parts of the system. Layout of devices, machines and workplaces was extensively studied during the last two decades (Gen & Cheng 1997). Proper placing of working devices (machines, technological equipment) within certain limits (e. g. workshop) is expected to reduce the cost of handling the materials, working means and auxiliary means (Tratnik, 2000). In most cases the minimum amount of cost of handling of work pieces is selected as the target function. In general those costs are the sum of the transport costs (which are proportional to the flow intensity and distances) and other costs. For the newly designed FMS the devices must be placed in best possible manner. The total variable transport costs of raw materials, work pieces and semi-finished products transported between the working devices within a certain time period are used as the quantitative criterion of placing of devices. First objective is to minimize material handling costs. Yet, other criteria like manpower requirements, work-in-process inventory, flow of information, etc. may play an important role, too. In all cases, certain departments need to be physically close to each other. Often one cannot satisfy all these needs by placing such departments next to each other, one has to decide which adjacencies are the most favourable (Dunker et. al. 2005). The optimum layout of devices and machines in the manufacturing system is one of the basic requirements for the design of the FMS, since good solutions in designing such system are prerequisite for its efficient functioning and low costs of operation. It has been estimated that 20% to 50% of the manufacturing costs are due to handling of work pieces; proper placing of devices can reduce the above costs for 10% to 30% (Tompkins et. al., 1997). Therefore the placing of devices must be known already in an early stage of designing of the system. This paper discusses the problem of placing of machines and devices in the FMS. The second part discusses the present state of research in this area. In the conclusion the discussion and the guidelines for further work follow.

2. PROBLEM OF LAYOUT OF DEVICES

The problem of layout of devices belongs to so-called problems of the NP completeness. The problems of NP exactingness are the problems which are not solvable within the polynomial time (Garey & Johnson 1979). The problem of the layout is of combinatory nature and is, therefore, suitable for solving by heuristic methods. The problem is theoretically solvable also by testing all possibilities (i.e. by blind searching), but practical experiences shows that in case of such manner of solving the capacities of either human or computer are quickly extended due to great a number of possible solutions. Thus, the principal problem in searching for the solution in case of combinatory problems is the control of the explosion of combinations.

2.1 Problem of designing of FMS

Most conventional approaches to placing of objects on the surface are based on the assumption that the positions are determined in advance. For placing they adhere to an exactly determined procedure. Heragu and Kusiak pointed out the fact that the FMS differ from the conventional production systems (Heragu & Kusiak 1988). The FMS include the devices and the machines which usually do not have identical dimensions and also the distances between the individual machines are not firmly determined (Heragu & Kusiak 1988) (Fig. 1). Therefore it is not possible to determine in advance the locations and then to place the devices on them. Due to those facts it is practically impossible in the FMS to locate the working device and workplaces by methods based on the principle "one place one device".

[FIGURE 1 OMITTED]

Therefore many heuristic procedures have been developed that also use different assumptions and simplifications, so the solution is only approximate. The target of all methods is the minimum transport costs, but they differ in exactingness, particularly in the length of the procedure. However, it cannot be decided with certainty which basic method and/or method of improvement of the layout is the best.

However, the area of researches is still always interesting for many researchers, since today the problems are solved by new, more modern methods and with the possibility of application of much greater computation capacity of modern computers. Today this area of researches covers different method and objectives. With respect to the objectives to be reached by placing of the devices also appropriateness of the methods of layout differs.

3. SURVEY OF STATE OF RESEARCHES

All methods have in common that they assume the space to be unchangeable, only the division of the space can be changed. The space can be represented with respect to the characteristics of the objects placed (Liggett, 2000):

* Space divided into separate parts (in this case of placing is effected on the principle: one object on one place).

* Space represented as free surface (in this case several objects are accommodated onto the common surface).

* Space represented as surface and form.

Computer aided layout of devices requires evaluation of the layout. There are three principal manners of evaluation of layout. According to the first manner the layout is evaluated by means of one cost function only. Usually, this is a cost function according to which the minimum of cost with respect to the cost of transport of material between devices is searched for. Although this cost function does not include many other criteria of layout this manner of evaluation is the most frequent. According to the second manner the layout is evaluated with respect to the relations between the neighbouring devices. This approach is based on the theory of graphs. The third manner of evaluation optimizes the layout with respect to several criteria by taking mutual dependencies into account. The position, orientation, immediate vicinity, path, distance etc. can be taken into account.

Layout of devices on the basic of the principle "one device on one place" is suitable for most conventional production systems. These methods include the assumption that there are n possible locations for n working devices so that it is possible to place exactly one device onto one location. The problem is treated as a quadratic assignment problem (Armour & Buffa 1963). The quadratic assignment problem belongs to problem of NP (non polynomial) completeness. For that reason different strategies of searching for high-quality solutions have been developed. In general, those strategies can be divided into three types (Liggett, 2000):

* Construction--opening strategies.

* Improvement strategies.

* Hybrid methods.

The opening strategies place the devices one after another onto their places by steps determined in advance and thus they make the final solutions. On the contrary, the improvement strategies start from the initial solutions and then they improve it step by step; two of such methods are the method of simulated annealing and the genetic algorithm method. The hybrid methods combine the structural and improvement strategies. For problems where the space is divided into equal separate parts the methods of layout assign one device to one location. However such approach is limited to and suitable only for solving some problems. Thus, such approach is not suitable for forming the FMS.

The space required by the individual devices in the FMS is of different size. Therefore the following approaches have been developed for placing the devices in the space treated as a unit:

* Modular method.

* Method of arrangement of rectangles.

* Method with limitations.

* Method on the basis of theory of graphs.

* Method with expert systems.

4. COMPUTER AIDED LAYING OUT OF DEVICES

Although the development of the computer aided placing has lasted already a long time, very few commercial products are available on the market. Though many CAD programmes have some capabilities of placing, in most cases they do not proceed beyond the graphic user interface for manual drawing of the layout. Almost no commercially available programme ensures the use of the problem information and of the optimization procedures.

The modern commercial programme must contain the algorithms for optimization of layout and the interactive user interface so that the human can influence the optimization process. The final layout and the technical documentation result from cooperation with the human.

5. PROSPECTS AND CONCLUSION

Due to the appearance of new knowledge's especially in the field of informatics and intelligent methods we are expecting a fresh impetus in this field.

The problem of searching for optimal design of FMS is theoretically solvable with testing of all possible solutions. That kind of search of solutions is the most general. One of the methods which use this kind of search is genetic algorithms. In case of designing of FMS, genetic algorithms methods prove to be very efficient. In the recent years the genetic algorithms have proved to be appropriate for solving such kind of problems in different areas of technology.

With genetic algorithms it is quite easily to form a target function which will include more economical and technical criteria and not only variable transport costs. Thus by introducing additional criteria by evaluation a more real target function would be obtained.

6. REFERENCES

Armour, G. C. & E. S. Buffa (1963). A heuristic algorithm and simulation approach to relative allocation of facilities. Management Science, Vol. 9, pp. 294-309, ISSN 0025-1909

Dunker, T,; Radons G. & Westkamper E. (2005). Combining evolutionary computation and dynamic programming for solving a dynamic facility layout problem. European Journal of Operational Research, Vol. 165, Issue 1, 16 August 2005, Pages 55-69, ISSN 0377-2217

Garey, M.R.; Johnson, D.S. (1979). Computers and Intractability: A Guide to the Theory of NP-Completeness, Freeman, ISBN 0-7167-1045-5, New York

Gen, M.; Cheng, R. (1997). Genetic Algorithms And Engineering Design, J. Wiley & Sons, Inc., ISBN 0-471-12741-8, New York

Heragu, S.; Kusiak, A. (1988). Machine layout problem in flexible manufacturing systems. Operations Research, Vol. 36, pp. 258-268, ISSN: 0030-364X

Liggett, R.S. (2000). Automated facilities layout: past, present and future. Automation in Construction, Vol. 9, pp. 197-215, ISSN 0926-5805

Tompkins, J. A.; White, J. A.; Bozer, Y.A.; Frazelle, E. H.; Tanchoco, J.M. & Trevino, J. (1996). Facilities Planning, John Wiley & Sons, ISBN 0-471-41389-5, New York

Tratnik, M. (2000). Razmescanje delovnih naprav pri delavniskem proizvodnem nacinu. Resitve proizvodnih problemov, pp. 179-187, Ljubljana, ISBN 961-6165-10-0