Annual Reviews in Control 31 (2007) 255267 www.elsevier.

com/locate/arcontrol

Facility layout problems: A survey

Amine Drira a,b,*, Henri Pierreval a, Sonia Hajri-Gabouj b
a

canique Avance e, Campus des Ce zeaux, BP 265, F-63175 Aubie ` re Cedex, France LIMOS, UMR CNRS 6158, IFMA, Institut Franc ais de Me b es et de Technologie, Centre Urbain Nord, BP 676, 1080 Tunis, Tunisia URAII, INSAT, Institut National des Sciences Applique Received 14 January 2007; accepted 4 April 2007 Available online 5 November 2007

Abstract Layout problems are found in several types of manufacturing systems. Typically, layout problems are related to the location of facilities (e.g., machines, departments) in a plant. They are known to greatly impact the system performance. Most of these problems are NP hard. Numerous research works related to facility layout have been published. A few literature reviews exist, but they are not recent or are restricted to certain specic aspects of these problems. The literature analysis given here is recent and not restricted to specic considerations about layout design. We suggest a general framework to analyze the literature and present existing works using such criteria as: the manufacturing system features, static/dynamic considerations, continual/discrete representation, problem formulation, and resolution approach. Several research directions are pointed out and discussed in our conclusion. # 2007 Elsevier Ltd. All rights reserved.
Keywords: Manufacturing facility; Facility layout; Material handling; Dynamic layout; Optimization methods

1. Introduction The placement of the facilities in the plant area, often referred to as facility layout problem, is known to have a signicant impact upon manufacturing costs, work in process, lead times and productivity. A good placement of facilities contributes to the overall efciency of operations and can reduce until 50% the total operating expenses (Tompkins et al., 1996). Simulation studies are often used to measure the benets and performance of given layouts (Aleisa & Lin, 2005). Unfortunately, layout problems are known to be complex and are generally NP-Hard (Garey & Johnson, 1979). As a consequence, a tremendous amount of research has been carried out in this area during the last decades. A few surveys have been published to review the different trends and research directions in this area. However, these surveys are either not recent (Hassan, 1994; Kusiak & Heragu, 1987; Levary & Kalchik, 1985), or focus on a very specic aspect of layout

design, such as loop layouts (Asef-Vaziri & Laporte, 2005), dynamic problems (Balakrishnan & Cheng, 1998) and design through evolutionary approaches (Pierreval, Caux, Paris, & Viguier, 2003). Benjaafar, Heragu, and Irani (2002) conducted a prospective analysis and suggested research directions. Our conclusion will show that several of their research propositions remain valid but other issues can also be raised. In this article, we present a recent survey about layout problems based on numerous literature references. First, in Section 2, we consider several possible denitions of layout problems. Then, we propose a general framework that can be used to analyze the current literature. Section 3 distinguishes the major features of the workshops that can be found. In Section 4, emphasis is put on so called dynamic problems. Section 5 discusses how facility layout problems can be formulated. In Section 6, we are interested in the approaches that are used to solve these problems. Although this review cannot be exhaustive, it has been conducted from a large number of literature references. 2. Denition of layout problems A facility layout is an arrangement of everything needed for production of goods or delivery of services. A facility is an entity that facilitates the performance of any job. It may be a

* Corresponding author at: URAII, INSAT, Institut National des Sciences es et de Technologie, Centre Urbain Nord, BP 676, 1080 Tunis, Applique Tunisia. Tel.: +216 71703829; fax: +216 71704329. E-mail addresses: Amine.Drira@ifma.fr (A. Drira), Henri.Pierreval@ifma.fr (H. Pierreval), Sonia.Gabouj@insat.rnu.tn (S. Hajri-Gabouj). 1367-5788/$ see front matter # 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.arcontrol.2007.04.001

256

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267

machine tool, a work centre, a manufacturing cell, a machine shop, a department, a warehouse, etc. (Heragu, 1997). Due to the variety of considerations found in the articles, researchers do not agree about a common and exact denition of layout problems. The most encountered formulations are related to static layout problems (in opposition to the dynamic layout problems that will be specically discussed in Section 3). Koopmans and Beckmann (1957) were among the rst to consider this class of problems, and they dened the facility layout problem as a common industrial problem in which the objective is to congure facilities, so as to minimize the cost of transporting materials between them. Meller, Narayanan, and Vance (1999) considered that the facility layout problem consists in nding a non-overlapping planar orthogonal arrangement of n rectangular facilities within a given rectangular plan site so as to minimize the distance based measure. Azadivar and Wang (2000) dened that the facility layout problem as the determination of the relative locations for, and allocation of, the available space among a given number of facilities. Lee and Lee (2002) reported that the facility layout problem consists in arranging n unequal-area facilities of different sizes within a given total space, which can be bounded to the length or width of site area in a way to minimize the total material handling cost and slack area cost. Shayan and Chittilappilly (2004) dened the facility layout problem as an optimization problem that tries to make layouts more efcient by taking into account various interactions between facilities and material handling systems while designing layouts. Numerous articles have been published in this area. In order to highlight what seems to constitute essential features to characterize layout problems, we propose in Fig. 1, a rst possible rough tree representation of the different factors taken into account in the literature. In fact, the problems addressed in research works differ, depending on such factors as: the workshop characteristics (e.g. ail, specicities of the manufacturing systems, the facility shapes, the material handling system, and the layout evolution), what is the problem addressed (e.g., problem formulation, objectives and constraints) and the approaches used to solve it (Resolution approaches). Although this tree representation can probably be improved in future research works, we have found it helpful in characterizing existing research works. Consequently, the rest of this article is organized in accordance with this representation and with the most important features identied. 3. Workshop characteristics impacting the layout Several types of workshop are addressed in the literature. In fact, the layout problems addressed are strongly dependent on the specic features of manufacturing systems studied. Several factors and design issues clearly differentiate the nature of the problems to be addressed, in particular: the production variety and volume, the material handling system chosen, the different possible ows allowed for parts, the number of oors on which the machines can be assigned, the facility shapes and the pickup and drop-off locations. Due to their importance, these factors are detailed below.

3.1. Products variety and volume The layout design generally depends on the products variety and the production volumes. Four types of organization are referred to in existing articles, namely xed product layout, process layout, product layout and cellular layout (Dilworth, 1996). These key organizations are sometimes discussed differently according to the authors. In Fixed product layout, the products generally circulate within the production facilities (machines, workers, etc.); in this particular type of layout, the product does not move, it is the different resources that are moved to perform the operations on the product. This type of layout is commonly found in industries that manufacture large size products, such as ships or aircrafts. Process layout groups facilities with similar functions together (resources of the same type). This organization is often reported to be suited when there is a wide variety of product. Product layout is used for systems with high production volumes and a low variety of products. Facilities are organized according to the sequence of the successive manufacturing operations. In Cellular layout, machines are grouped into cells, to process families of similar parts. These cells also need to be placed on the factory oor. Therefore, one is also generally concerned with so called intra cells machine layout problems, as mentioned for example in (Proth, 1992, ch. 3) and (Hamann & Vernadat, 1992). Here, one is concerned with nding the best arrangement of machines in each cell. 3.2. Facility shapes and dimensions Two different facility shapes are often distinguished (Fig. 2): regular, i.e., generally rectangular (Kim & Kim, 2000) and irregular, i.e., generally polygons containing at least a 2708 angle (Lee & Kim, 2000). As mentioned by Chwif, Pereira Barretto, and Moscato (1998) a facility can have given dimensions, dened by a xed length (Li) and a xed width (Wi). In this case, the facilities are called xed or rigid blocks. According to the same authors, a facility can also be dened by its area, its aspect ratio: ai = Li/Wi, an upper bound aiu and a lower bound ail such that ail  ai  aiu. The aspect ratio was also used by Meller et al. (1999). If ai = ail = aiu, this corresponds to the xed shape blocks case (Chwif et al., 1998). 3.3. Material handling systems A material handling system ensures the delivery of material to the appropriate locations. Material handling equipment can be conveyors (belt, roller, wheel), automated guided vehicles (AGV), robots, etc. (El-Baz, 2004). Tompkins et al. (1996) estimated that 2050% of the manufacturing costs are due to the handling of parts and then a good arrangement of handling devices might to reduce them for 1030%. When dealing with a material handling system, the problem consists in arranging facilities along the material handling path. Two dependent design problems are considered: nding the facility layout and selecting the handling equipment. The type of material-handling device determines the pattern to be used

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267

257

Fig. 1. Tree representation of the layout problems.

258

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267

Fig. 2. Regular and irregular facility shapes.

Fig. 3. Layout design considering material handling devices.

Fig. 4. Multi-oor layout.

for the layout of machine (Devise & Pierreval, 2000; Heragu & Kusiak, 1988). Co, Wu, & Reisman (1989) also pointed out that the facility layout impacts the selection of the handling device. Given the difculty of solving both problems jointly, they are mainly solved sequentially (Hassan, 1994). Among the major types of layout arrangement based on the type of material handling, one can distinguish, as depicted in Fig. 3: single row layout, multi-rows layout, loop layout and open-eld layout (Yang, Peters, & Tu, 2005). The single row layout problem occurs when facilities have to be placed along a line (Djellab & Gourgand, 2001; Ficko, Brezocnick, & Balic, 2004; Kim, Kim, & Bobbie, 1996; Kumar, Hadjinicola, & Lin, 1995). Several shapes may be considered from this basic situation, such as straight line, semicircular or U-shape (Hassan, 1994). The loop layout problem deals with the assignment of m facilities to candidate locations 1, . . ., m, in a closed ring network, around which parts are transported in one direction (Chaieb, 2002; Cheng & Gen, 1998; Cheng, Gen, & Tosawa, 1996; Nearchou, 2006; Potts & Whitehead, 2001). The loop layout incorporates a Load/Unload (L/U) station, i.e., location from which a part enters and leaves the loop. This station is unique and it is assumed to be located between position m and 1. The multi-rows layout involves several rows of facilities (Hassan, 1994). The movements of parts occur between facilities from the same row and from different rows (Chen, Wang, & Chen, 2001; Ficko et al., 2004; Kim et al., 1996). The open eld layout corresponds to situations where facilities can be placed without the restrictions or constraints that would be induced by such arrangements as single row or loop layout (Yang et al., 2005). 3.4. Multi-oor layout Nowadays, when it comes to construct a factory in urban area, land supply is generally insufcient and expensive. The limitation of available horizontal space creates a need to use a vertical dimension of the workshop. Then, it can be relevant to locate the facilities on several oors, as depicted in Fig. 4. This

gure shows that parts can move horizontally on a given oor (horizontal ow direction), but also from one oor to another oors located at a different level (vertical ow direction). The vertical movement of parts requires a vertical transportation device: elevator. In such situations, both the position on the oor and the levels have to be determined for each facility, so that the related problems are referred to as multi-oor layout problems (Kochhar & Heragu, 1998). Johnson (1982) seems to be among the rsts to address a multiple-oor layout problem. He dealt with the problem of dening relative locations of facilities in a multiple-oor building. Later, other researchers focused on taking into consideration vertical movements of parts from one oor to another (Bozer, Meller, & Erlebacher, 1994; Meller & Bozer, 1996, 1997). Elevators are often the material handling system reported (Lee, Roh, & Jeong, 2005). Their number and location are either known (Lee et al., 2005) or to be determined through optimization (Matsuzaki, Takashi, & Yoshimoto, 1999). In (Matsuzaki et al., 1999), the capacity of each elevator was considered as a constraint. The number of oors can be known (Lee et al., 2005) or to be determined, depending on each oor area and on the number and dimensions of the facilities (Patsiatzis & Papageorgiou, 2002). 3.5. Backtracking and bypassing Backtracking and bypassing (see Fig. 5) are two particular movements that can occur in ow-line layouts, which impact the ow of the products. Backtracking is the movement of a part, from one facility to another preceding it in the sequence of facilities in the ow-line arrangement (Braglia, 1996; Kouvelis

Fig. 5. Backtracking and bypassing.

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267

259

& Chiang, 1992; Zhou, 1998). The number of these movements has to be minimized. Zhou (1998) called this problem Production Line Formation Problem (PLFP), which consists in determining the orders (partial or total) of machines so as to minimize the weighted sum of arrows whose direction is contrary to the global ow of products, while taking into account constraints on the rank of machines. Bypassing occurs when a part skips some facilities during its moving towards the ow line arrangement (Chen et al., 2001). Hassan (1994) noticed that several procedures were presented for dealing with and minimizing backtracking but no procedure was suggested in the literature for addressing bypassing. 3.6. Pick-up and drop-off locations It is often necessary to determine the location from which parts enter and leave facilities, called Pick-up and Drop-off (P/ D) points. Although they can potentially be located at various places (Kim & Kim, 2000), several researchers restricted their possible position to reduce the complexity (Das, 1993; Rajasekharan, Peters, & Yang, 1998; Welgama & Gibson, 1993). An example is given in Fig. 6. 4. Static vs. dynamic layout problems We have seen that the workshop characteristics introduce differences in the way to design the layout. In addition, it is well known that nowadays, manufacturing plants must be able to respond quickly to changes in demand, production volume and product mix. Page (1991) reported that, on average, 40% of a companys sales come from new products. However, the change in product mix yields to modify the production ow and thus affects the layout. Gupta and Seifoddini (1990) stated that 1/3 of USA companies undergo major reorganization of the production facilities every 2 years. A good number of authors have tried to take such an important issue into account when designing the layout. Most articles dealing with layout problems are implicitly considered as static; in other words they assume that the key data about the workshop and what it is intended to produce will remain constant enough over a long period of time. Recently the idea of dynamic layout problems has been introduced by several researchers. Dynamic layout problems take into account possible changes in the material handling ow over multiple periods (Balakrishnan, Cheng, Conway, & Lau, 2003; Braglia, Zanoni, & Zavanella, 2003; Kouvelis, Kurawarwala, & Gutierrez, 1992; Meng, Heragu, & Zijm, 2004). In this respect, the planning horizon is generally divided into periods that may be dened in weeks, months, or years. For each period, the estimated ow data remains constant. A layout plan for the dynamic layout problem consists of series of layouts, each layout being associated with a period.

Fig. 7. Evolution of the layout over four periods.

Fig. 7 shows a layout with six equal size locations to be arranged in each of the four periods in the planning horizon. The objective can be to determine a layout for each period in the planning horizon, while minimizing the sum of the material handling costs, for all periods, and the sum of the rearrangement costs between time periods (Balakrishnan, Cheng, Conway, et al., 2003; Baykasoglu, Dereli, & Sabuncu, 2006). Rearrangement costs have to be considered when facilities need to be moved from one location to another (Baykasoglu & Gindy, 2001). 5. Formulation of layout problems The workshop characteristics and the static or dynamic issues being raised, there are several ways of formulating mathematically the layout problems so that they can be solved. This formulation of static and dynamic layout problems can be based on several types of models, which allow the complex relationships between the different elements involved in a layout problem to be expressed. Such models can rely on different principles, which include graph theory (Kim & Kim, 1995; Leung, 1992; Proth, 1992) or neural network (Tsuchiya, Bharitkar, & Takefuji, 1996). These models are generally used to suggest solutions to the layout problems, which most researchers consider as optimization problems, with either single or multiple objectives. Depending on the manner in which the problem is formulated, that is, discrete or continuous, the formulations found in the literature can lead to Quadratic Assignment Problems (QAP) or Mixed Integer Programmings (MIP), which are the most commonly encountered. In each case, a few authors have argued that the available data could not be perfectly known and have suggested fuzzy formulation. These approaches are discussed in the following. 5.1. Discrete formulation The layout is sometimes considered as discrete (Fig. 8a). In such a case, the associated optimization problem is sometimes addressed as QAP. The plant site is divided into rectangular blocks with the same area and shape, and each block is assigned

Fig. 6. Example of P/D points of a machine with a regular shape.

Fig. 8. Discrete and continual layout representations.

260

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267

to a facility (Fruggiero, Lambiase, & Negri, 2006). If facilities have unequal areas, they can occupy different blocks (Wang, Hu, & Ku, 2005). A typical formulation, when determining the relative locations of facilities so as to minimize the total material handling cost, is as follows (Balakrishnan, Cheng, & Wong, 2003): min s.t.
N X i 1 N X j 1 N X N X N X N X i1 j1 k1 l1

5.2. Continual formulation In many articles, the layout representation is continual (Fig. 8b). It is often addressed as Mixed Integer Programming Problems (Das, 1993). All the facilities are placed anywhere within the planar site and must not overlap each other (Das, 1993; Dunker et al., 2005; Meller et al., 1999). The facilities in the plant site are located either by their centroid coordinates (xi,yi), half length li and half width wi or by the coordinates of bottom-left corner, length Li and width Wi of the facility. The distance between two facilities can be, for example, expressed through the rectilinear norm (Chwif et al., 1998): d i j xi ; yi ; x j ; y j jxi x j j jyi y j j (4)

f ik d jl X i j X kl

(1)

X i j 1;

j 1; . . . ; N

(2)

X i j 1;

i 1; . . . ; N

(3)

where N is the number of facilities in the layout, f ik the ow cost from facility i to k, djl the distance from location j to l and Xij the 0, 1 variable for locating facility i at location j. The objective function (1) represents the sum of the ow costs over every pair of facilities. Eq. (2) ensures that each location contains only one facility and Eq. (3) guarantees that each facility is placed only in one location. Discrete formulations are suggested, for example, by Kouvelis and Chiang (1992) and Braglia (1996) to minimize part backtrack in single row layouts. The same type of approach is also used by Afentakis (1989), to design a loop layout, so as to minimize the trafc congestion, i.e., the number of times a part traverses the loop before all its operations are completed. There are two kinds of congestion measures commonly used in loop layout design: Min-Sum and Min-Max. A Min-Sum problem attempts to minimize the total congestion of all parts; while a Min-Max problem attempts to minimize the maximum congestion among a family of parts (Cheng & Gen, 1998; Cheng et al., 1996; Nearchou, 2006). Discrete representation of the layout is commonly used for dynamic layout problems. The problems addressed are related to equal size facilities (Baykasoglu & Gindy, 2001; Lacksonen & Enscore, 1993) and must respect constraints ensuring that each location is assigned to only one facility at each period, and that exactly one facility is assigned to each location at each period (Baykasoglu & Gindy, 2001; McKendall, Shang, & Kuppusamy, 2006). Budget constraints can be added to carry out the reconguration of facilities on the oor plant (Balakrishnan, Robert Jacobs, & Venkataramanan, 1992; Baykasoglu et al., 2006). In fact, the rearrangement costs must not exceed a certain level of the budget. Discrete representations are not suited to represent the exact position of facilities in the plant site and can not model appropriately specic constraints as the orientation of facilities, pick-up and drop-off points or clearance between facilities. In such cases, a continuous representation is found to be more relevant by several authors (Das, 1993; Dunker, Radonsb, & mpera, 2005; Lacksonen, 1997). Westka

The pick-up and drop-off points can generate constraints in the layout problem formulation (Kim & Kim, 2000; Welgama & Gibson, 1993; Yang et al., 2005). In this case, the distance traveled by a part from the drop-off of facility i to the pick-up of facility j, can for example, be given by Eq. (5) (Kim & Kim, 2000).
I O I d i j jxO i x j j jyi y j j

(5)

O where (xO i ; yi ) designate the coordinates of the drop-off point of facility i, and (xIj ; yIj ) the coordinates of the pick-up point of facility j. The determination of the best locations of the P/D stations is a specic problem, addressed for example by Chittratanawat and Noble (1999), Kim and Kim (1999), and Aiello, Enea, and Galante (2002). Obviously, area constraints on the plant site exist, which require the total area available to be superior or equal to the sum of all the facility areas. The area allocated to each machine on the oor plan must also take into account the space of other resources or buffers, which are needed to operate the machine (Lacksonen, 1997). The clearance between facilities can be included or not in the facility surface (Braglia, 1996; Heragu & Kusiak, 1988, 1991). Another very important constraint is that facilities must not overlap. Welgama and Gibson (1993) set two conditions for the non-overlapping of facilities: condition of X-projection nonoverlapping and condition of Y-projection non-overlapping:

x jt xib x jb xit  0 y jt yib y jb yit  0

(6) (7)

where (xit, yit) and (xib, yib) are the top-left and the bottom right corners of the facility i and (xjt, yjt) and (xjb, yjb) are the top-left and the bottom right corners of the facility j. Mir and Imam (2001) dened an overlap area Aij between two facilities to formulate this constraint. The layout optimization problem is expressed as follows: Minimize objective function Subject to Ai j  0 (8)

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267

261

where Ai j li j DX i j DY i j ;   Li L j DX i j li j jxi x j j; 2   Wi W j DY i j li j jyi y j j; 2  1 for DX i j  0 and DY i j  0 li j 1 otherwise Li ; W i are the length and width of facility i, and (xi,yi) are coordinates of facility i. Other constraints can also be considered in the layout formulation, such as a pre-dened orientation of certain facilities (Dunker et al., 2005). Given such constraints, a typical formulation of the optimization problem can be as follows: Minimize C
N X N X i1 j1 I O f i j jxIj xO i j jy j yi j

(9)

where N is the number of facilities, f ij the amount/cost of material ow from drop-off point of facility i to pick-up O point of facility j, (xO i ; yi ) the coordinates of drop-off point of I I facility i, and (x j ; y j ) are the coordinates of pick-up point of facility j. Very few works seem to deal with dynamic layout problems with a continuous representation. Dunker et al. (2005) addressed unequal size layout problems in a dynamic environment and assumed that the facility sizes vary from one period to another. 5.3. Fuzzy formulation In many concrete cases, data affecting layout problems are not exactly known. Stochastic approaches, such as the use of queuing networks (Meng et al., 2004) are seldom seen. Fuzzy logic has been proposed to handle the imprecision or uncertainty that is often encountered (Evans, Wilhlem, & Karwowsky, 1987; Grobelny, 1987a; Raoot & Rakshit, 1991). A few approaches based on fuzzy concepts exist to design layouts. Evans et al. (1987) addressed the placement of unequal size facilities on the plant area. They expressed relations between every pair of facilities by fuzzy relations describing closeness and importance. These relations allow the analyst to specify the importance associated with each pair of facilities to be located at any distance from each other. The authors proposed a fuzzy formulation of the problem through linguistic variables and propose a heuristic. Grobelny (1987a, 1987b) tacked the problem of locating n facilities to n xed locations so as to minimize the total material handling cost. The data impacting the layout, such as closeness links and trafc intensity, are fuzzy and modeled with linguistic variables and fuzzy implications. A heuristic procedure, based on binary fuzzy relations, is developed for the selection and the placement of facilities in the available locations. Several principles of this approach are also used by Raoot and Rakshit (1991), who considered the problem of nding the best arrangement on

the plant site of facilities based on specications about their inter-relationship, which are characterized through linguistic variables. Gen, Ida, and Cheng (1995) addressed a multiobjective multi-rows layout problem with unequal area. They are interested in situations where the clearance cannot be precisely dened, and is therefore considered as fuzzy. In Dweiri and Meier (1996), who dealt with a discrete facility layout problem, the amount of parts circulating between facilities, the amount of communications between facilities (information ow) and the number of material handling equipments used to transfer parts between facilities are considered as fuzzy factors. The authors developed an Activity Relationship Chart (ARC) based on the judgment of experts that is used to specify relationships between each pair of facilities. ARC is then integrated in the well known heuristic CORELAP to nd the best placement of facilities. Aiello and Enea (2001) argued that the product market demands are uncertain data that can be dened as fuzzy numbers. They minimize the total material handling cost, along a single row conguration, under the constraints that the capacity of production for each department is limited. To solve a single row layout problem, they split the fuzzy demands in a-cuts and determine the a-level fuzzy cost for each possible layout. Deb and Bhattacharyya (2005), addressed the placement of facilities with pick-up and drop-off points in a continual plane, so as to minimize the total material handling cost. The position of facilities depends on such factors as: the personal ow, the supervision relationships, the environmental relationships and the information relationships, which are rated using linguistic variables (e.g., high, medium, low). The authors developed a fuzzy decision support system based on a set of fuzzy IFTHEN rules. A construction heuristic is then used do determine the placement of facilities in the plant site. 5.4. Multi-objective layout problems In most articles about layout problems, the main objective is to minimize a function related to the travel of parts (the total material handling cost, the travel time, the travel distance, etc.). To be more realistic, some researchers have considered more than a single objective. For example, Dweiri and Meier (1996) aimed at minimizing simultaneously the material handling ow and the equipment ow and the information ow. Most authors combine the different objectives into a single one either by means of Analytic Hierarchy Process (AHP) methodology (Harmonosky & Tothero, 1992; Yang & Kuo, 2003) or using a linear combination of the different objectives (Chen & Sha, 2005). Few researchers used a Pareto approach to generate a set of non-dominated solutions. Aiello, Enea, and Galante (2006) dealt with a layout problem related to the minimization of the material handling cost and the maximization of an adjacency function (assessment of the proximity requests between two departments). The set of non-dominated solutions is then found and a best solution is then selected from this set using the well known Electre method.

262

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267

5.5. Simultaneous solving of different problems It is common that other problems have to be solved together with the layout design. For example, this occurs when designing cellular manufacturing systems, where one has both to assign machines to cells (cell formation problems) and to determine the position of each machine in the cell (intra cell layout). The position of each cell in the oor plant has also to be determined. Instead of formulating and solving these problems sequentially, it is sometimes possible to address these two issues as a same problem (Gupta, Gupta, Kumar, & Sundaram, 1996). 6. Resolution approaches Several approaches exist to address the different types of problems that are formulated in the literature. They aim either at nding good solutions, which satises certain constraints given by the decision maker or at searching for an global or local optimum solutions given one or several performance objectives. This has yield heuristic based methods or optimization algorithms, as explained in the following of this section. Some attempts of using articial intelligent approaches have been made to address layout problems. Expert systems were, for example, proposed in (Heragu & Kusiak, 1990). Hamann and Vernadat (1992) also used this approach for intra-cell problems. More recently, an expert system based on articial neural networks was implemented for facility layout construction in a manufacturing system (Chung, 1999). Several types of optimization approaches have been proposed in the literature. In the following, we distinguish: exact methods, such as branch and bound, and approximated approaches, such as heuristics and metaheuristics. 6.1. Exact approaches Among articles that dealt with exact methods, Kouvelis and Kim (1992) developed a branch and bound algorithm for the unidirectional loop layout problem. Meller et al. (1999) also used this approach to solve the problem of placing n rectangular facilities within a given rectangular available area. They proposed general classes of valid inequalities, based on an acyclic sub-graph structure, to increase the range of solvable problems and use them in a branch-and-bound algorithm. Kim and Kim (1999) addressed the problem of nding P/D locations on xed size facilities for a given layout. The objective of the problem is to minimize the total distance of material ows between the P/D points. Authors suggested a branch and bound algorithm to nd an optimal location of the P/D points of each facility. Rosenblatt (1986) used a dynamic programming method to solve a dynamic layout problem with equal size facilities. However, only small problem instances have been solved optimally (six facilities and ve time periods).

6.2. Approximated approaches Since exact approaches are often found not to be suited for large size problems, numerous researchers have developed heuristics and metaheuristics. Construction approaches build progressively the sequence of the facilities until the complete layout is obtained whereas improvement methods start from one initial solution and they try to improve the solution with producing new solution. Construction heuristics include: CORELAP (Lee & Moore, 1967), ALDEP (Seehof & Evans, 1967) and COFAD (Tompkins & Reed, 1976), SHAPE (Hassan, Hogg, & Smith, 1986). Example of improvement heuristics are: CRAFT (Armour & Buffa, 1963), FRAT (Khalil, 1973) and DISCON (Drezner, 1987). Among the approaches based on metaheuristics, one can distinguish global search methods (Tabu search and simulated annealing) and evolutionary approaches (genetic and ant colony algorithms). Chiang and Kouvelis (1996) developed a tabu search algorithm to solve a facility layout problem. They used a neighborhood based on the exchange of two locations of facilities and included a long term memory structure, a dynamic tabu list size, an intensication criteria and diversication strategies. Chwif et al. (1998) used a simulated annealing algorithm to solve the layout problem with aspect ratio facilities sizes. Two neighborhood procedures are proposed: a pairwise exchange between facilities and random moves on the planar site in the four main directions (upwards, downwards, leftwards and rightward). McKendall et al. (2006) suggested two simulated annealing approaches for a dynamic layout problem with equal size facilities. The rst simulated annealing approach used a neighborhood based on a descent pairwise exchange method, which consists in randomly changing the location of two facilities while the solution is improved. The second approach combines the rst simulated annealing algorithm and improvement strategy called look-ahead and look-back strategy. Genetic algorithms seem to become quite popular in solving facility layout problems (Pierreval et al., 2003). In fact, a large number of studies using such approaches have been published: see Banerjee and Zhou (1995), Mak, Wong, & Chan (1998), Tam and Chan (1998), Azadivar and Wang (2000), Wu and mpera (2003), Appleton (2002), Dunker, Radonsb, & Westka and Wang et al. (2005) for the static layout problems, and Balakrishnan and Cheng (2000), Balakrishnan, Cheng, Conway, et al. (2003), and Dunker et al. (2005) for the dynamic layout problems. A very important problem when developing a genetic algorithm is related to the coding of a candidate oor plan. A popular representation of the continual layout is the slicing tree (Shayan & Chittilappilly, 2004). A slicing tree is composed of internal nodes partitioning the oor plan and of external nodes representing the facilities. Each internal node can be labeled either h (horizontal) or v (vertical), indicating whether it is a horizontal or vertical slice whereas external nodes label the facility index (1, 2, 3, . . ., n for n facilities). Each rectangular partition corresponds to a space allocated to a facility. Fig. 9 shows a particular layout and the corresponding slicing tree.

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267

263

Fig. 9. Slicing tree representation of the oor plan.

Wu and Appleton (2002) suggested a slicing tree to represent simultaneously the layout and the aisles and adapted genetic operators. From a given layout, the slicing tree is generally encoded into a string form, in order to use particular genetic operators. Tam (1992) suggested coding a solution by a binary string with two parts, which represent operators and operands, and latter in three parts: the tree structure, the operators and the operands (Tam & Chan, 1998). Al-Hakim (2000) improved Tam Chans approach (1998) and proposed a new operator named transplanting, to ensure the coherence of an offspring. The problem of avoiding reparation procedures when dealing with slicing trees is tackled by Shayan and Chittilappilly (2004). When authors addressed discrete layout problems, the coding scheme differs from continual representation. For discrete representation, a popular solution for coding layouts is based on Space Filling Curves (SFC) (Wang et al., 2005). The plant area being divided into grids, a space lling curve denes a continuous sequence through all neighbored squares in the underlying layout (Fig. 10). Space-lling curves ensure that a facility is never split (Bock & Hoberg, 2007). Nevertheless, this technique requires many rules to verify the connection of all positions of a layout as for example using expert rules (Wang et al., 2005). When a space lling strategy is dened, solutions have to be coded. Islier (1998) decomposed strings into three segments, encoding the sequence of facilities, the area required for each facility and the width of each sweeping band. Recently, Wang et al. (2005) encoded the chromosomes genes through ve segments strings. The rst segment shows the department

placement sequence. The second gives the required areas of each department. The third segment indicates the site size (length and width). The fourth segment shows the sweeping direction (1: horizontal, 2: vertical) and the fth segment indicates the sweeping bands. An example is illustrated in Fig. 10. The objective function used in evolutionary methods is generally expressed as a mathematical cost function, which is derived from the problem formulation under consideration. To take into account in a more realistic way the system performance, simulation models have been connected to evolutionary methods to evaluate the candidate solutions (Azadivar & Wang, 2000). Hamamoto, Yih, and Salvendy (1999) addressed a real problem of pharmaceutical industry. The chromosome evaluation is performed through the simulation of a 4 months production. Ant colony optimization has been recently applied for solving layout problems. Solimanpur, Vrat, and Shankar (2005) developed an ant algorithm for a sequence-dependent single row machine layout problem. Baykasoglu et al. (2006) proposed an ant colony algorithm for solving the unconstrained and budget constrained dynamic layout problems. The hybridization of different metaheuristics has also been considered for solving facility layout problems. Mahdi, Amet, and Portman (1998) proposed a hybrid approach for minimizing the material handling cost. They used a simulated annealing algorithm to solve the geometrical aspect of the problem, a genetic algorithm to make decisions about the material handling system and an exact method (Hitchcocks method) to minimize the total material handling utilization cost. Mir and Imam (2001) presented a hybrid approach for a layout problem with unequal area facilities. Starting from an initial solution given by a simulated annealing algorithm, the optimal positions of facilities are determined by an analytical search technique in a multi-stage optimization process. Lee and Lee (2002) presented a hybrid genetic algorithm for a xed shape and unequal area facility layout problem. Tabu search and simulated annealing are rst used to nd global solutions and the genetic algorithm is introduced in the middle of the local search process to search for a global solution. Balakrishnan, Cheng, Conway, et al. (2003) developed a hybrid genetic algorithm to solve the dynamic layout problem previously tackled by Rosenblatt (1986). The initial population is generated with two methods: a random method and an Urbans procedure (Urban, 1993). The crossover is based on a dynamic programming approach and the mutation is achieved by the CRAFT heuristic (Armour & Buffa, 1963). McKendall and Shang (2006) developed and compare three hybrid ant colony algorithms for a dynamic facility layout problem. They combine an ant colony with three local search procedures: (1) a random descent pairwise exchange procedure, (2) a simulated annealing algorithm and (3) a look-ahead/look-back procedure. 7. Conclusion and research directions In this article, we have presented a recent comprehensive survey related to facility layout problems. Although this survey cannot be exhaustive, the analysis carried out is based on a large

Fig. 10. String scheme of a discrete layout representation based on a space lling curve (Wang et al., 2005).

264

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267

number of literature references. From this analysis, it appears that articles related to layout design continue to be regularly published in major research journals and that facility layout remains an open research issue. Several current trends and directions seem worth being mentioned. First of all, recent papers include more and more complex and/or realistic characteristics of the studied manufacturing systems. Typical examples are P/D points, aisles, complex geometrical constraints, several oors, which are taken into account together when formulating the layout design problem. This is indeed an important issue because many articles contain restrictive assumptions that are not adapted to the complexity of many manufacturing system facilities. This is an old trend (Benjaafar et al., 2002). However, research is still needed. The use of a third dimension when designing a plant is a recent consideration that certainly requires more research, for example to select and optimize resources related to the vertical transportation of parts between different oors. The often unrealistic aspect of static approaches, which consider that the data available are relevant to characterize the future operating conditions of the system, seems to be now well identied by researchers. Dynamic approaches may sometimes be a potential alternative; meanwhile, they often rely on knowledge of the future operating conditions. Fuzzy methods can offer interesting possibilities to include uncertainty. However, as already noted by Benjaafar et al. (2002), research is still needed to suggest or improve methods for designing (1) robust and adaptive layouts, (2) sensitivity measures and analysis of layouts and (3) stochastic models used to evaluate solutions. In terms of methods used to solve layout problems, one can obviously see that the use of metaheuristics is more and more reported in articles, in order to cope with problems of a larger size and to take into account more realistic constraints. Evolutionary algorithms seem to be among the most popular approaches. Solving methods are also hybridized, either to solve complex problems (e.g., metaheuristics embedding heuristics or connected with exact methods) or to provide more realistic solutions (e.g., connection of evolutionary principles with simulation). Approaches based on articial intelligence are now seldom published. Given the fact that it is probably difcult to solve everything without using some kind of expert knowledge about the system, there is probably still a need for hybrid methods capable of optimizing the layout while taking into account expert available knowledge. We have noticed that most published works focus on determining the position of facilities. However, in practice this problem is often consider together with other design problems, such as the choice of the type of manufacturing or transportation resources, the design of cells, the determination of resources capacities, etc. These problems are often not independent (for example, the choice of a conveyor as a material handling device does not induce the same constraints as the choice of automated guided vehicles). Therefore, there is still research needed for solving the different problems involved in the design of the workshop simultaneously instead of sequentially. Such combinations, that start to be addressed, turn

out to be promising and would worth being developed and improved. This would lead to favor more global research about workshop design, instead of concentrating on facility layout problems. The articles studied in this article focus on manufacturing system applications. However, layout design problems also concern other types of systems, such as ports, supermarkets, airports, etc. These constitute interesting areas that could benet from the advances made in the specic area of manufacturing. As a nal remark, let us note that commercial software tools available on the market to globally assist in the design of manufacturing are currently limited. Therefore, there is probably a need for trying to make the resolution approaches more generic, so that they can be embed as layout procedures in software tools supporting the design of manufacturing systems. The combination with graphical tools would also render such tools more efcient and attractive and a few authors have started to consider possible interfaces with virtual reality systems. References
Afentakis, P. A. (1989). Loop layout design problem for exible manufacturing systems. International Journal of Flexible Manufacturing Systems, 1, 143 175. Aiello, G., & Enea, M. (2001). Fuzzy approach to the robust facility layout in uncertain production environments. International Journal of Production Research, 39(18), 40894101. Aiello, G., Enea, M., & Galante, G. (2002). An integrated approach to the facilities and material handling system design. International Journal of Production Research, 40(15), 40074017. Aiello, G., Enea, M., & Galante, G. (2006). Multi-objective approach to facility layout problem by genetic search algorithm and Electre method. Robotics and Computer-Integrated Manufacturing, 22, 447455. Al-Hakim, L. (2000). On solving facility layout problems using genetic algorithms. International Journal of Production Research, 38(11), 2573 2582. Aleisa, E. E., & Lin, L. (2005). For effectiveness facilities planning: Layout optimization then simulation, or vice versa? In Proceedings of the 2005 Winter Simulation Conference. Armour, G. C., & Buffa, E. S. (1963). A heuristic algorithm and simulation approach to relative allocation of facilities. Management Science, 9(2), 294 300. Asef-Vaziri, A., & Laporte, G. (2005). Loop based facility planning and material handling. European Journal of Operational Research, 164(1), 111. Azadivar, F., & Wang, J. (2000). Facility layout optimization using simulation and genetic algorithms. International Journal of Production Research, 38(17), 43694383. Balakrishnan, J., & Cheng, C. H. (1998). Dynamic layout algorithms: A stateof-the-art survey. Omega, 26(4), 507521. Balakrishnan, J., & Cheng, C. H. (2000). Genetic search and the dynamic layout problem. Computers & Operations Research, 27(6), 587593. Balakrishnan, J., Cheng, C. H., & Wong, K. F. (2003a). FACOPT: A user friendly FACility layout OPTimization system. Computers & Operations Research, 30(11), 16251641. Balakrishnan, J., Cheng, C. H., Conway, D. G., & Lau, C. M. (2003b). A hybrid genetic algorithm for the dynamic plant layout problem. International Journal of Production Economics, 86(2), 107120. Balakrishnan, J., Robert Jacobs, F., & Venkataramanan, M. A. (1992). Solutions for the constrained dynamic facility layout problem. European Journal of Operational Research, 57(2), 280286.

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267 Banerjee, P., & Zhou, Y. (1995). Facilities layout design optimization with single loop material ow path conguration. International Journal of Production Research, 33(1), 183204. Baykasoglu, A., Dereli, T., & Sabuncu, I. (2006). An ant colony algorithm for solving budget constrained and unconstrained dynamic facility layout problems. Omega, 34(4), 385396. Baykasoglu, A., & Gindy, N. N. Z. (2001). A simulated annealing algorithm for dynamic layout problem. Computers & Operations Research, 28(14), 1403 1426. Benjaafar, S., Heragu, S. S., & Irani, S. A. (2002). Next generation factory layouts: Research challenges and recent progress. Interface, 32(6), 5876. Bock, S., & Hoberg, K. (2007). Detailed layout planning for irregularly-shaped machines with transportation path design. European Journal of Operational Research, 177, 693718. Bozer, Y. A., Meller, R. D., & Erlebacher, S. J. (1994). An improvement-type layout algorithm for single and multiple oor facilities. Management Science, 40(7), 918932. Braglia, M. (1996). Optimization of a simulated-annealing-based heuristic for single row machine layout problem by genetic algorithm. International Transactions in Operational Research, 3(1), 3749. Braglia, M., Zanoni, S., & Zavanella, L. (2003). Layout design in dynamic environments: Strategies and quantitative indices. International Journal of Production Research, 41(5), 9951016. ` mes de production Chaieb, I. (2002). Conception et exploitation des syste ` res: Introduction des ta ches de transport. Ph.D. exibles manufacturie cialite en productique automatique dissertation (in French). France: Spe et informatique industrielle, Ecole centrale de Lille. Chen, C. W., & Sha, D. Y. (2005). Heuristic approach for solving the multiobjective facility layout problem. International Journal of Production Research, 43(21), 44934507. Chen, D. S., Wang, Q., & Chen, H. C. (2001). Linear sequencing for machine layouts by a modied simulated annealing. International Journal of Production Research, 39(8), 17211732. Cheng, R., & Gen, M. (1998). Loop layout design problem in exible manufacturing systems using genetic algorithms. Computers & Industrial Engineering, 34(1), 5361. Cheng, R., Gen, M., & Tosawa, T. (1996). Genetic algorithms for designing loop layout manufacturing systems. Computers & Industrial Engineering, 31(3 4), 587591. Chiang, W. C., & Kouvelis, P. (1996). An improved tabu search heuristic for solving facility layout design problems. International Journal of Production Research, 34(9), 25652585. Chittratanawat, S., & Noble, J. S. (1999). An integrated approach for facility layout, P/D location and material handling system design. International Journal of Production Research, 37(3), 683706. Chung, Y. K. (1999). A neuro-based expert system for facility layout construction. Journal of Intelligent Manufacturing, 10(5), 359385. Chwif, L., Pereira Barretto, M. R., & Moscato, L. A. (1998). A solution to the facility layout problem using simulated annealing. Computers in Industry, 36(12), 125132. Co, H. C., Wu, A., & Reisman, A. (1989). A throughput-maximizing facility planning and layout model. International Journal of Production Research, 27(1), 112. Das, S. K. (1993). A facility layout method for exible manufacturing systems. International Journal of Production Research, 31(2), 279297. Deb, S. K., & Bhattacharyya, B. (2005). Fuzzy decision support systems for manufacturing facilities layout planning. Decision Support Systems, 40, 305314. Devise, O., & Pierreval, A. (2000). Indicators for measuring performances of morphology and materials handling systems. International Journal of Production Economics, 64(13), 209218. Dilworth, J. B. (1996). Operation management. McGraw Hill. Djellab, H., & Gourgand, A. (2001). A new heuristic procedure for the singlerow facility layout problem. International Journal of Computer Integrated Manufacturing, 14(3), 270280. Drezner, Z. (1987). A heuristic procedure for the layout of a large number of facilities. International Journal of Management Science, 33(7), 907 915.

265

mpera, E. (2003). A coevolutionary algoDunker, T., Radonsb, G., & Westka rithm for a facility layout problem. International Journal of Production Research, 41(15), 34793500. mpera, E. (2005). Combining evolutionary Dunker, T., Radonsb, G., & Westka computation and dynamic programming for solving a dynamic facility layout problem. European Journal of Operational Research, 165(1), 5569. Dweiri, F., & Meier, F. A. (1996). Application of fuzzy decision-making in facilities layout planning. International Journal of Production Research, 34(11), 32073225. El-Baz, M. A. (2004). A genetic algorithm for facility layout problems of different manufacturing environments. Computers & Industrial Engineering, 47(23), 233246. Evans, G. W., Wilhlem, M. R., & Karwowsky, W. (1987). A layout design heuristic employing the theory of fuzzy sets. International Journal of Production Research, 25, 14311450. Ficko, M., Brezocnick, M., & Balic, J. (2004). Designing the layout of singleand multiple-rows exible manufacturing system by genetic algorithms. Journal of Materials Processing Technology, 157158: 150158. Fruggiero, F., Lambiase, A., & Negri, F. (2006). Design and optimization of a facility layout problem in virtual environment.. In Proceeding of ICAD 2006 (pp. 2206). Garey, M. R., & Johnson, D. S. (1979). Computers and intractability: A guide to the theory of NP-completeness. New York: WH Freeman. Gen, M., Ida, K., & Cheng, C. (1995). Multi row machine layout problem in fuzzy environment using genetic algorithms. Computers & Industrial Engineering, 29(14), 519523. Grobeiny, J. (1987a). The fuzzy approach to facility layout problems. Fuzzy Sets and Systems, 23, 175190. Grobelny, J. (1987b). On one possible fuzzy approach to facility layout problems. International Journal of Production Research, 25, 11231141. Gupta, T., & Seifoddini, H. (1990). Production data based similarity coefcient for machinecomponent grouping decisions in the design of cellular manufacturing system. International Journal of Production Research, 28(4), 12471269. Gupta, Y. P., Gupta, M. C., Kumar, A., & Sundaram, C. (1996). A genetic algorithm-based approach to cell composition and layout design problems. International Journal of Production Research, 34(2), 447482. Hamamoto, S., Yih, Y., & Salvendy, G. (1999). Development and validation of genetic algorithm-based facility layout-a case study in the pharmaceutical industry. International Journal of Production Research, 37, 749768. Hamann, T., & Vernadat, F. (1992). The intra cell layout problem in automated manufacturing system. 8th international Conference on CAD/CAM, robotics and factory of the future (CARs & FOF 92). Harmonosky, C. M., & Tothero, G. K. (1992). A multi-factor plant layout methodology. International Journal of Production Research, 30(8), 1773 1789. Hassan, M. M. D. (1994). Machine layout problem in modern manufacturing facilities. International Journal of Production Research, 32(11), 2559 2584. Hassan, M. M. D., Hogg, G. L., & Smith, D. R. (1986). SHAPE: A construction algorithm for area placement evaluation. International Journal of Production Research, 24(5), 12831295. Heragu, S. S. (1997). Facilities design. Boston: BWS. Heragu, S. S., & Kusiak, A. (1988). Machine layout problem in exible manufacturing systems. Operations Research, 36(2), 258268. Heragu, S. S., & Kusiak, A. (1990). Machine layout: An optimization and knowledge-based approach. International Journal of Production Research, 28, 615635. Heragu, S. S., & Kusiak, A. (1991). Efcient models for the facility layout problem. European Journal of Operational Research, 53(1), 113. Islier, A. A. (1998). A genetic algorithm approach for multiple criteria facility layout design. International Journal of Production Research, 36(6), 1549 1569. Johnson, R. V. (1982). SPACECRAFT for multi-oor layout planning. Management Sciences, 28(4), 407417. Khalil, T. M. (1973). Facilities relative allocation technique (FRAT). International Journal of Productions Research, 11(2), 183194.

266

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267 Meller, R. D., Narayanan, V., & Vance, P. H. (1999). Optimal facility layout design. Operations Research Letters, 23(35), 117127. Meng, G., Heragu, S. S., & Zijm, H. (2004). Recongurable layout problem. International Journal of Production Research, 42(22), 47094729. Mir, M., & Imam, M. H. (2001). A hybrid optimization approach for layout design of unequal-area facilities. Computers & Industrial Engineering, 39(12), 4963. Nearchou, A. C. (2006). Meta-heuristics from nature for the loop layout design problem. International Journal of Production Economics, 101(2), 312328. Page, A. L. (1991). New product development survey: Performance, and best practices. PDMA Conference. Patsiatzis, D. I., & Papageorgiou, L. G. (2002). Optimal multi-oor process plant layout. Computers and Chemical Engineering, 26(45), 575583. Pierreval, H., Caux, C., Paris, J. L., & Viguier, F. (2003). Evolutionary approaches to the design and organization of manufacturing systems. Computers & Industrial Engineering, 44(3), 339364. Potts, C. N., & Whitehead, J. D. (2001). Workload balancing and loop layout in the deign of a exible manufacturing system. European Journal of Operational Research, 129(2), 326336. ` mes de production. Presses Proth, J. M. (1992). Conception et gestion des syste Universitaires de France. pp. 6877. Rajasekharan, M., Peters, B. A., & Yang, T. (1998). A genetic algorithm for facility layout design in exible manufacturing systems. International Journal of Production Research, 36(1), 95110. Raoot, A. D., & Rakshit, A. (1991). A fuzzy approach to facilities layout planning. International Journal of Production Research, 29, 835857. Rosenblatt, M. J. (1986). The dynamics of plant layout. Management Science, 32(1), 7686. Seehof, J. M., & Evans, W. O. (1967). Automated layout design program. The Journal of Industrial Engineering, 18, 690695. Shayan, E., & Chittilappilly, A. (2004). Genetic algorithm for facilities layout problems based on slicing tree structure. International Journal of Production Research, 42(19), 40554067. Solimanpur, M., Vrat, P., & Shankar, R. (2005). An ant algorithm for the single row layout problem in exible manufacturing systems. Computers & Operations Research, 32(3), 583598. Tam, K. Y. (1992). Genetic algorithms, function optimization and facility layout design. European Journal of Operational Research, 63(2), 322346. Tam, K. Y., & Chan, S. K. (1998). Solving facility layout problems with geometric constraints using parallel genetic algorithms: Experimentation and ndings. International Journal of Production Research, 36(12), 3253 3272. Tompkins, J. A., & Reed, J. R. (1976). An applied model for the facilities design problem. International Journal of Production Research, 14, 583595. Tompkins, J. A., White, J. A., Bozer, Y. A., Frazelle, E. H., Tanchoco, J. M., & Trevino, J. (1996). Facilities planning. New York: Wiley. Tsuchiya, K., Bharitkar, S., & Takefuji, Y. (1996). A neural network approach to facility layout problems. European Journal of Operational Research, 89(3), 556563. Urban, T. L. (1993). A heuristic for the dynamic facility layout problem. IIE Transactions, 25(4), 5763. Wang, M. J., Hu, M. H., & Ku, M. H. (2005). A solution to the unequal area facilities layout problem by genetic algorithm. Computers in Industry, 56(2), 207220. Welgama, P. S., & Gibson, P. R. (1993). A construction algorithm for the machine layout problem with xed pick-up and drop-off points. International Journal of Production Research, 31(11), 25752590. Wu, Y., & Appleton, E. (2002). The optimisation of block layout and aisle structure by a genetic algorithm. Computers & Industrial Engineering, 41(4), 371387. Yang, T., & Kuo, C. (2003). A hierarchical AHP/DEA methodology for the facilities layout design problem. European Journal of Operational Research, 147, 128136. Yang, T., Peters, B. A., & Tu, M. (2005). Layout design for exible manufacturing systems considering single-loop directional ow patterns. European Journal of Operational Research, 164(2), 440455.

Kim, C. B., Kim, S. S., & Bobbie, L. F. (1996). Assignment problems in singlerow and double-row machine layouts during slow and peak periods. Computers & Industrial Engineering, 30(3), 411422. Kim, J. G., & Kim, Y. D. (1999). A branch and bound algorithm for locating input and output points of departments on the block layout. Journal of the operational research society, 50(5), 517525. Kim, J. G., & Kim, Y. D. (2000). Layout planning for facilities with xed shapes and input and output points. International Journal of Production Research, 38(18), 46354653. Kim, J. Y., & Kim, Y. D. (1995). Graph theoretic heuristics for unequal-sized facility layout problems. Omega, 23(4), 391401. Kochhar, J. S., & Heragu, S. S. (1998). MULTI-HOPE: A tool for multiple oor layout problems. International Journal of Production Research, 36(12), 34213435. Koopmans, T. C., & Beckmann, M. (1957). Assignment problems and the location of economic activities. Econometrica, 25(1), 5376. Kouvelis, P., & Chiang, W. C. (1992). A simulated annealing procedure for the single row layout problems in exible manufacturing systems. International Journal of Production Research, 30, 717732. Kouvelis, P., & Kim, M. W. (1992). Unidirectional loop network layout problem in automated manufacturing systems. Operations Research, 40, 533550. Kouvelis, P., Kurawarwala, A. A., & Gutierrez, G. J. (1992). Algorithms for robust single and multiple period layout planning for manufacturing systems. European Journal of Operations Research, 63(2), 287303. Kumar, K. R., Hadjinicola, G. C., & Lin, T. L. (1995). A heuristic procedure for the single-row facility layout problem. European Journal of Operational Research, 87(1), 6573. Kusiak, A., & Heragu, S. S. (1987). The facilities layout problem. European Journal of Operational Research, 29(3), 229251. Lacksonen, T. A. (1997). Preprocessing for static and dynamic facility layout problems. International Journal of Production Research, 35(4), 10951106. Lacksonen, T. A., & Enscore, E. E. (1993). Quadratic assignment algorithms for the dynamic layout problem. International Journal of Production Research, 31, 503517. Lee, G. C., & Kim, Y. D. (2000). Algorithms for adjusting shapes of departments in block layouts on the gird-based plane. Omega, 28(1), 111122. Lee, K. Y., Roh, M. I., & Jeong, H. S. (2005). An improved genetic algorithm for multi-oor facility layout problems having inner structure walls and passages. Computers & Operations Research, 32(4), 879899. Lee, R., & Moore, J. M. (1967). CORELAP-computerized relationship layout planning. The Journal of Industrial Engineering, 18, 195200. Lee, Y. H., & Lee, M. H. (2002). A shape-based block layout approach to facility layout problems using hybrid genetic algorithm. Computers & Industrial Engineering, 42, 237248. Leung, J. (1992). A graph-theoretic heuristic for exible manufacturing systems. European Journal of Operational Research, 57(2), 243252. Levary, R. R., & Kalchik, S. (1985). Facilities layout-a survey of solution procedures. Computers & Industrial Engineering, 9(2), 141148. Mahdi, A. H., Amet, H., & Portman, M. C. (1998). Physical layout with minimization of the transport cost (Research Internal Report). Nancy, France: LORIA. Mak, K. L., Wong, Y. S., & Chan, F. T. S. (1998). A genetic algorithm for facility layout problems. Computer Integrated Manufacturing Systems, 11(12), 113127. Matsuzaki, K., Takashi, I., & Yoshimoto, K. (1999). Heuristic algorithm to solve the multi-oor layout problem with the consideration of elevator utilization. Computers & Industrial Engineering, 36(2), 487502. McKendall, A. R., & Shang, J. (2006). Hybrid ant systems for the dynamic facility layout problem. Computers & Operations Research, 33(3), 790803. McKendall, A. R., Shang, J., & Kuppusamy, S. (2006). Simulated annealing heuristics for the dynamic facility layout problem. Computers & Operations Research, 33(8), 24312444. Meller, R. D., & Bozer, Y. A. (1996). A new simulated annealing algorithm for the facility layout problem. International Journal of Production Research, 34, 16751692. Meller, R. D., & Bozer, Y. A. (1997). Alternative approaches to solve the multioor facility layout problem. Journal of Manufacturing Systems, 16(3), 192203.

A. Drira et al. / Annual Reviews in Control 31 (2007) 255267 ` Zhou, J. (1998). Algorithmes et outils pour lanalyse des ux de production a laide du concept dordre. Ph.D. dissertation (in French). University of Strasbourg 1. Amine Drira is a doctoral student in industrial computing at the Institute of Applied Sciences and Technology of Tunis, Tunisia (University of 7 November at Carthage). He is currently working as an assistant professor at the High Institute of Medical Technologies (University Tunis El Manar). His research interests include facility layout, fuzzy computation, metaheuristics and multiobjective optimization. Henri Pierreval is a professor at the French Institute of Mechanical Engineering (IFMA) of Clermont-Ferrand in France. His research activities are carried out within the LIMOS laboratory of Clermont-Ferrand (Laboratory of Computing, System Modeling and Optimization): UMR CNRS 6158. He received his doctoral degree and qualication to direct research degree from the University of Lyon in France. The design, modeling, operation of manufacturing

267

systems and of networks of rms are the main focuses of Dr. Pierrevals research. He is the author or co-author of numerous articles in these areas, published in well-known international refereed journals. He participates in national and international research collaborations and is on the editorial board of scientic journals. Sonia Hajri-Gabouj is an associate professor at the Institute of Applied Sciences and Technology of Tunis, Tunisia. She received her BS degree in electrical engineering from the National Engineering School of Monastir, Tunisia, in 1994. She got her MS and PhD degrees in industrial computing and automatic from the University of Sciences and Technologies of Lille, France, in 1994 and 1997, respectively. She received a qualication to direct research degree from the National Engineering School of Tunis in 2003. She is co-responsible for the URAII research unit in automatic and industrial computing of the Institute of Applied Sciences and Technology of Tunis. Her research interests include fuzzy modeling, design and control of manufacturing systems, scheduling and optimization and metaheuristics.