Multisystem of Material Handling for Shipyard Facility Layout Optimization Using NSGA-Ⅱ
https://doi.org/10.1007/s11804-025-00643-2
-
Abstract
The need to transport goods across countries and islands has resulted in a high demand for commercial vessels. Owing to such trends, shipyards must efficiently produce ships to reduce production costs. Layout and material flow are among the crucial aspects determining the efficiency of the production at a shipyard. This paper presents the initial design optimization of a shipyard layout using Nondominated Sorting Algorithm-Ⅱ (NSGA-Ⅱ) to find the optimal configuration of workstations in a shipyard layout. The proposed method focuses on simultaneously minimizing two material handling costs, namely work-based material handling and duration-based material handling. NSGA-Ⅱ determines the order of workstations in the shipyard layout. The semiflexible bay structure is then used in the workstation placement process from the sequence formed in NSGA-Ⅱ into a complete design. Considering that this study is a case of multiobjective optimization, the performance for both objectives at each iteration is presented in a 3D graph. Results indicate that after 500 iterations, the optimal configuration yields a work-based MHC of 163 670.0 WBM-units and a duration-based MHC of 34 750 DBM-units. Starting from a random solution, the efficiency of NSGA-Ⅱ demonstrates significant improvements, achieving a 50.19% reduction in work-based MHC and a 48.58% reduction in duration-based MHC.Article Highlights● The paper optimizes shipyard layouts using NSGA-Ⅱ, aiming to minimize both work-based and duration-based material handling costs.● The research introduces duration-based material handling costs (DBM) as an additional objective, addressing the gap in previous studies that only considered volume and distance without accounting for material delivery time.● The study tackles unequal-area shipyard layouts, optimizing for both material transport distance and delivery time to enhance overall efficiency. -
1 Introduction
Facility layout planning (FLP) is the process of producing optimum facility layout configurations to achieve efficiency in some operations, such as the production of goods and public services. In naval engineering, FLP can be applied to design shipyard layouts for efficient ship production. In general, the process and flow of ship production are still carried out based on the experience of ship designers. With FLP, the ship production flow can be simulated by a computer through a systematic layout design using various methods and algorithms.
Researchers have adopted FLP to find the optimum facility layout. Liu and Liu (2019) conducted a literature survey of FLP and found that 25 out of 26 papers on FLP aimed to minimize material handling costs (MHCs). Braglia et al. (2003) stated that excellent facility layout design can increase material handling efficiency between facilities and reduce MHC by approximately 10%–30%. Matson et al. (1992) stated that MHC contributes around 30%–40% of total production costs and can increase to 70% in some industry types. Some of the widely used research methods for FLP are heuristic algorithms, such as genetic algorithms (Besbes et al., 2020; 2021; Deep, 2020; Gonçalves and Resende, 2015; Kochhar et al., 1998; Mak et al., 1998; Palomo-Romero et al., 2017; Wang et al., 2005), ant colony algorithm (Hani et al., 2007; Komarudin and Wong, 2010; Kulturel-Konak and Konak, 2011; Liu and Liu, 2019; Wong and Komarudin, 2010; Yu-Hsin Chen, 2013; Zouein and Kattan, 2022), simulated annealing (Allahyari and Azab, 2018; Palubeckis, 2017; Sulaiman et al., 2021; Turgay, 2018), and particle swarm algorithm (Guan et al., 2019; Liu et al., 2018; Muhayat and Utamima, 2024; Samarghandi et al., 2010). To study the application of FLP in shipyard layout, Choi et al. (2017) carried out a two-step optimization consisting of topology design and geometry design. For topology design, a genetic algorithm is used to obtain the optimum layout while minimizing MHCs. The results are then converted into a geometry design using a stochastic growth algorithm. The ultimate goal is to achieve alignment conditions in departments that must be close together. Junior et al. (2023) expounded on this study by adding the ELECTRE method and local search to the stochastic growth algorithm for geometry design. If some departments still do not meet the proximity constraint, then the local search engine is executed. Gunawan et al. (2021) used a genetic algorithm to minimize the MHC of 25 topological departments, several of which have fixed positions. Türk et al. (2022) compared the performance of several types of operators and methods in genetic algorithms to produce the most effective technique in determining topology designs for shipyards.
The fundamental aim of a material handling system is to efficiently transport the appropriate blend of tools and materials (including raw materials, components, and finished products) to the correct location, in the proper state orientation, and precisely on schedule while minimizing the overall cost (Rosenblatt, 2013). According to this explanation, MHC can be defined in various ways depending on priorities: volume of material transported, duration of material delivery, and location between facilities. In the shipyard layout optimization studies conducted by some of researchers (Choi et al., 2017; Gunawan et al., 2021; Junior et al., 2023; Junior and Azzolini, 2019), the MHC calculation only considers the volume of material and distance between departments. The duration of material delivery (weighted distance calculation) is not taken into account. Therefore, the current work addresses this gap by adding material delivery duration as an objective. The two objectives will be "work-based" MHC, which considers the volume of material and distance between departments, and "duration-based" MHC, which considers the duration of materials delivery between departments. Although the actual layout of ship production workshops is a continuous case, research on shipyard layout optimization using a heuristic approach introduces a novel methodology that allows for the discretization of the problem. Taking into account the unique size of each workshop during the optimization, the present work employs a topological approach for shipyards with unequal areas. Previous research using a topological approach assumes that workshops are of equal size throughout the optimization. This work aims to address this gap by employing a topological approach for unequal areas.
2 Problem and methods
2.1 Problem description
The data and conditions of the shipyard in this study are displayed in Tables 1‒4. Table 1 provides information on the size of each department in the shipyard, and Table 2 gives a classification of these departments according to their functionality. Table 3 provides information regarding material flow between departments, and Table 4 provides information on transporters or equipment used to send materials between departments.
Table 1 Information on buildings and facilitiesNo. Name Size (m) Position category 1 Profile stockyard 8×4 Free 2 Straightening area 15×3 Free 3 Cutting area 7×9 Free 4 Bending area 4×4 Free 5 Piping workshop 7×5 Free 6 Mechanical workshop 6×6 Free 7 Electrical workshop 4×4 Free 8 Part assembly 7×6 Free 9 Sub-assembly 10×7 Free 10 Panel production area 5×6 Free 11 Block assembly 8×8 Free 12 First pre-erection 8×5 Free 13 Second pre-erection 17×9 Free 14 Pre-outfitting 5×4 Free 15 Paint workshop 8×5 Free 16 Warehouse 7×7 Free 17 Stock space (first quay) 4×16 Fix 18 Stock space (second quay) 4×16 Fix 19 Free space area 6×6 Fix 20 Waste material area 7×6 Free 21 Fire protection facilities 6×6 Fire 22 Office 5×6 Fix 23 Refreshing room & Toilet 5×4 Fix 24 Parking area 6×6 Fix 25 Entrance area 5×5 Fix Table 2 Group of workshopsFabrication workshop Assembly yard Paint workshop Berths Additional areas 1 8 15 17 16 2 9 18 20 3 10 19 21 4 11 22 5 12 23 6 13 24 7 14 25 Table 3 Material flows between departmentsNo. From To Quantity (t) 1 1 2 1 300 2 2 3 1 100 3 2 4 200 4 3 4 1 020 5 4 8 850 6 4 9 180 7 15 13 1 180 8 8 9 680 9 8 11 130 10 11 15 2 700 11 11 14 550 12 11 12 700 13 5 14 550 14 14 13 620 15 13 17 900 16 13 18 900 Table 4 Specification of material transporterNo. Speed (m/min) Capacity (t) 1 12 50 2 12 50 In Tables 3 and 4, "t" means tons. In Table 1, the position of each department is defined by position category. If a department has the status "Free", then it can be anywhere in the equal area layout. However, if a department has the status "Fix", then it must be in a particular square in the topological layout. Figure 1 depicts the initial conditions of the topological layout in this study.
2.2 Comparison method
Table 5 compares methods and problems in this work with those of Choi et al. (2017) and Junior et al. (2023). The fixed position department already has a geometrically fixed position. In other words, the total area and shape used for the shipyard have been regulated, along with several departments that must be in certain positions. The remaining empty land will become the area of the optimized departments.
Table 5 Comparison method and problem of related papersConsideration Fixed Position Status MHC Factor Type of Distance Department conditions when MHC was calculated Choi et al. (2017) Topologically And Geometrically Work Based (MV & DD) Euclidian Equal Area Junior et al. (2023) Topologically and Geometrically Work Based (MV & DD) Euclidian Equal Area Gunawan et al. (2021) Topologically (equal area) Work Based (MV & DD) Euclidian Equal Area This Paper Topologically (equal area) Work Based (MV & DD), Duration Based (TD) Manhattan Unequal Area Notations: MV: Volume of materials that have to be transported, DD: Distance between departments, TD: Duration of material transport process In this study, the total area used for the shipyard is not determined in advance, and the departments that must be in a fixed position are also only within the topological design review. For example, Figure 1 shows that department 17 must be in the top proper position (square 25). Its coordinates in the geometric design depend on the placement results from the proposed method. Hence, its natural position has yet to be determined at the start.
The following table shows the coordinates of the center points of several fixed departments based on Figure 3. Figure 3 further explains that the geometric position of the fixed department can vary depending on the solution produced, in this case, the solution by the algorithm.
Figure 2 Initial layout of the shipyard (example from Choi et al. (2017), Junior et al. (2013))2.3 Methodology
The configuration of the shipyard layout is based on the sequence of genes (departments) within the NSGA-11 individual. After the individual algorithms have been fully formed, the sequence is converted into a sequence in the shipyard layout. Given that the shipyard layout is a continuous problem, the solution is limited to the factorial value of the number of existing departments and can go to infinity. A suitable approach is warranted to simplify the process of finding the optimum solution to this continuous problem. This study uses NSGA-Ⅱ as an algorithm to produce department sequences. These sequences are then converted into a shipyard layout using the horizontal layer method, also called a semiflexible bay structure. After the layout has been formed, the fitness values, namely "work-based" MHC and "duration-based" MHC, are calculated. In this study, the shipyard's layout is considered a 2D building, namely, from the top view.
2.3.1 NSGA-Ⅱ
First introduced in 2002 by Kalyanmoy Deb, NSGA-Ⅱ is an optimization algorithm for multiobjective cases. Similar to the genetic algorithm, NSGA-Ⅱ uses the principles of natural selection among living creatures for the survival of the fittest. In essence, NSGA-Ⅱ works differently in each case. The following is the pseudocode of NSGA-Ⅱ, which is also used in this study.
2.3.2 Layout design
The departments in the shipyard layout are grouped into several bays or layers. In the flexible bay structure (FBS), each bay does not need to have the same width and number of departments. The width of each bay is automatically adjusted by the number of departments it contains. In addition, the bays in FBS can be depicted vertically or horizontally. In this study, the semiflexible bay structure method is used, namely, a modification of FBS with each bay having the same number of departments.
In this study, a separation model is used in horizontal bays or horizontal layers. Each bay contains five departments, and the distance between adjacent departments is 1 m. Each department has four sides, and each bay has four sides. Figures 5 and 6 provide an overview of the representation of departments and bays.
Algorithm 1 Pseudocode from NSGA-Ⅱ START Desired number of generations = n_total Generation = n Starting with n = 1 Generating an initial population (Pn) of x individuals Conduct evaluations on the initial population As long as n ≤ n_total: Creating an offspring population of x individuals through the process of crossover and mutation Combining the initial population (Pn) and the derived population (Qn) into a combined population (Rn) Perform ranking (nondominated sorting) on Rn into several fronts (F) Take x/2 top individuals in Rn to become the next population (Pn+1): If the (x/2+1)th and (x/2−1)th individuals are not in the same rank: Continue If the (x/2+1)th and (x/2−1)th individuals are in the same rank (Fz): Perform crowding distance procedures on Fz and sort them Selecting some individuals at Fz amounts to a loss of Pn+1 Pn+1 has been fulfilled n plus 1 FINISH Parameters and variables:
n = Total number of bays = 25
ni = Total number of departments in bay i = 5
SNSGA−Ⅱ=x1, x2, x3, x4, …, xn
Kj = The sets of departments in bay j
rx = Closest side distance of 2 departments horizontally in the same bay = 1
ry = Distance between layer = 1
Dj, z= Department in bay j sequence z
Wbi = Width of bay i
Lbi = Length of bay i
Wdi = Width of department i
Ldi = Length of department i
Cxj, z= Centroid x of department in bay j in order to z
Cyj, z= Centroid y of department in bay j in order to z
Models:
$$ K_j=\left\{a \in S_{\mathrm{NSGA}-\mathrm{Ⅱ}} \mid\left[(j-1) \cdot n_j\right]+1 \leqslant a \leqslant j \cdot n_j\right\} $$ (1) $$ \mathrm{Wb}_i=\max\nolimits _{a=1}^{n_i} \mathrm{Wd}_a $$ (2) $$ \mathrm{Lb}_i=\left[\sum\nolimits_{a=1}^{n_i} \mathrm{Ld}_a\right]+\left(n_i-1\right) \cdot \mathrm{rx} $$ (3) If j = 1:
$$ C y_{j, z}=\mathrm{Wb}_j-\left(0, 5 \cdot \mathrm{Wd}_{D_{j, z}}\right) $$ (4) $$ C x_{j, z}=\left[\sum\nolimits_{a=1}^z \operatorname{Ld}_{D_{j, a}}\right]+(z-1) \cdot \mathrm{rx}-\left(0, 5 \cdot \operatorname{Ld}_{D_{j, z}}\right) $$ (5) If j > 1:
$$ C y_{j, z}=\left[\sum\nolimits_{a=1}^{j-1} \mathrm{~Wb}_a\right]+(j-1) \cdot \mathrm{ry}+\left(0, 5 \cdot \mathrm{Wd}_{D_{j, z}}\right) $$ (6) $$ C x_{j, z}=\left[\sum\nolimits_{a=1}^z \operatorname{Ld}_{D_{j, a}}\right]+(z-1) \cdot \mathrm{rx}-\left(0, 5 \cdot \operatorname{Ld}_{D_{j, z}}\right) $$ (7) According to the equations above, for bay 1, the S1 part of the department intersects with the S1 part of the bay. For the other bays, the S3 section of the department overlaps with the S3 section of the bay. This placement method also provides evidence that the width and length of each bay constantly change with each modification in the solution of the algorithm. Ultimately, the departments with fixed positions will have geometric places that are also continually evolving.
2.3.3 Calculation of objectives
Choi et al. (2017), Junior et al. (2023), Gunawan et al. (2021), Junior and Azzolini (2019) performed MHC calculations when the departments were still in equal areas; that is, all departments have the same size. For the present study, the calculation of the two objectives, namely "work-based" MHC and "duration-based" MHC, is carried out when the departments are already in an unequal area and considered to have different sizes (Table 1). The Manhattan distance is used to measure the distance between two departments. For the first objective, namely "work-based" MHC or "weighted distance, " only the volume of transported material and the delivery distance are considered. For the second objective, namely "duration-based" MHC, the following diagram explains the scheme (suppose the transporter will deliver materials from workshop x to r).
• Each transporter can only send material according to its capacity on each trip (Table 4).
• Each transporter can run at the same time.
• The trips considered in duration-based MHC are only those on the red line, specifically when carrying material loads.
• The blue line and dashed blue line represent idle capacity, meaning the transporter is not carrying any load. These trips are not included in duration-based MHC.
The following section explains the calculations of the objectives used in the mathematical model.
Parameters and variables:
dmab = Manhattan distance between Department a and Department b
fa, b = Material flow from department a to department b (given in Table 2)
Nx = number of transporters = 2
vi = speed of transporter i (given in Table 3)
Ci= Capacity of transporter i (t)
Wa, b = Amount of materials that needs to be transported from dept a to dept b (t)
Ta, b = The time required to trip from dept a to dept b
Ya, b = Total time required to trip from dept a to dept b when
Pa, b = Number of trips required from dept a to dept b
Models:
$$ r 1_{a, b}=\left|C y_a-C y_b\right| $$ (8) $$ r 2_{a, b}=\left|C x_a-C x_b\right| $$ (9) $$ \mathrm{dm}_{a, b}=\left|r 1_{a b}+r 2_{a b}\right| $$ (10) $$ T_{a, b}=\frac{\mathrm{dm}_{a, b}}{v_i} $$ (11) $$ P_{a, b}=\left\lceil\frac{W_{a, b}}{N_x \cdot C_i}\right\rceil $$ (12) $$ Y_{a, b}=T_{a, b} \cdot P_{a, b} $$ (13) $$ T_{a, b}=\frac{\mathrm{dm}_{a, b}}{v_i} $$ (14) Objectives:
$$ \begin{gathered} \operatorname{Minimize}\left[\sum\limits_a \sum\limits_b f_{a, b} \cdot \mathrm{dm}_{a b}\right] \\ \text { Minimize }\left[\sum\limits_{k=1}^n Y_{i_k, j_k}\right] \end{gathered} $$ k represents the index of the row in the Table 2, ${i_k}~\mathop {and}\limits_{.~.~.} $ jk denote the workshops 'From' and 'To' in the kth row, respectively.
This model aims to optimize material transportation between departments by minimizing the total distance and travel time. According to Equations (8) and (9), the Manhattan distance is calculated by summing the absolute differences in the x and y coordinates. Equation (10) combines these results to obtain dmab. Equation (11) calculates the travel time (Ta, b) based on the distance and the transporter's speed. Equation (12) determines the number of trips (Pa, b) required based on the transporter's capacity and the amount of material. Equation (13) then calculates the total travel time (Ya, b) by multiplying the travel time by the number of trips. The first objective is to minimize the total distance by multiplying the material flow with the Manhattan distance and summing it for all department pairs. The second objective is to minimize the total travel time by summing the total time required for all trips. With this model, we can achieve operational efficiency in material transportation between departments.
3 Optimization results
Optimization is carried out using the Python 3 programming language. NSGA-Ⅱ has the following characteristics:
1) Performs 500 generations/iterations
2) Has 30 individuals in each population (Pn)
3) The crossover process uses a cyclic operator.
4) In the mutation process, random methods are taken at each iteration, namely, the swap, flip, and slide methods.
For multiobjective optimization, each iteration can have multiple optimal solutions, which are collectively part of the Pareto optimal solutions. However, the solution points from each iteration depicted in Figure 9 are selected based on the highest crowding distances.
4 Analysis and discussion
This study fundamentally differs from similar papers by considering the duration of material delivery between departments. The optimal layout of the shipyard should not only involve weighted distance but should also consider the time of content delivery. Figure 9 shows that the algorithm model converges relatively quickly at around 500 iterations. This finding indicates that the developed algorithm model is capable of reaching an optimal solution in a short amount of time. The figure also shows that the work-based MHC and duration-based MHC are approaching their respective minimum values throughout the iteration.
Table 7 Characteristics of best solutions in the last iterationSequence of departments "Work Based" MHC "Duration Based" MHC [25, 24, 23, 1, 16, 22, 9, 2, 6, 21, 20, 8, 4, 3, 7, 19, 12, 11, 15, 10, 18, 5, 14, 13, 17] 163 670 WBM-unit 34 750 DBM-unit 5 Conclusions
This research presents an optimization of shipyard layout configuration by considering the weighted distance and total duration of material delivery between departments for each production cycle using NSGA-Ⅱ. This study aims to simultaneously minimize weighted distance and the duration of material delivery between departments to maximize production efficiency in the shipyard. Results showed that NSGA-Ⅱ solves this multiobjective problem while providing optimum results.
The following conclusions can be drawn from this research: 1) NSGA-Ⅱ can solve multiobjective problems for production efficiency in shipyard layout configuration; 2) The optimal results obtained after 500 iterations show that the work-based MHC reaches 163 670.0 WBM-units, and the duration-based MHC reaches 34.750 DBM-units; 3) Starting from a random solution, the efficiency of NSGA-Ⅱ is 50.19% for work-based MHC and 48.58% for duration-based MHC.
Competing interestThe authors have no competing interests to declare that are relevant to the content of this article. -
Figure 2 Initial layout of the shipyard (example from Choi et al. (2017), Junior et al. (2013))
Table 1 Information on buildings and facilities
No. Name Size (m) Position category 1 Profile stockyard 8×4 Free 2 Straightening area 15×3 Free 3 Cutting area 7×9 Free 4 Bending area 4×4 Free 5 Piping workshop 7×5 Free 6 Mechanical workshop 6×6 Free 7 Electrical workshop 4×4 Free 8 Part assembly 7×6 Free 9 Sub-assembly 10×7 Free 10 Panel production area 5×6 Free 11 Block assembly 8×8 Free 12 First pre-erection 8×5 Free 13 Second pre-erection 17×9 Free 14 Pre-outfitting 5×4 Free 15 Paint workshop 8×5 Free 16 Warehouse 7×7 Free 17 Stock space (first quay) 4×16 Fix 18 Stock space (second quay) 4×16 Fix 19 Free space area 6×6 Fix 20 Waste material area 7×6 Free 21 Fire protection facilities 6×6 Fire 22 Office 5×6 Fix 23 Refreshing room & Toilet 5×4 Fix 24 Parking area 6×6 Fix 25 Entrance area 5×5 Fix Table 2 Group of workshops
Fabrication workshop Assembly yard Paint workshop Berths Additional areas 1 8 15 17 16 2 9 18 20 3 10 19 21 4 11 22 5 12 23 6 13 24 7 14 25 Table 3 Material flows between departments
No. From To Quantity (t) 1 1 2 1 300 2 2 3 1 100 3 2 4 200 4 3 4 1 020 5 4 8 850 6 4 9 180 7 15 13 1 180 8 8 9 680 9 8 11 130 10 11 15 2 700 11 11 14 550 12 11 12 700 13 5 14 550 14 14 13 620 15 13 17 900 16 13 18 900 Table 4 Specification of material transporter
No. Speed (m/min) Capacity (t) 1 12 50 2 12 50 Table 5 Comparison method and problem of related papers
Consideration Fixed Position Status MHC Factor Type of Distance Department conditions when MHC was calculated Choi et al. (2017) Topologically And Geometrically Work Based (MV & DD) Euclidian Equal Area Junior et al. (2023) Topologically and Geometrically Work Based (MV & DD) Euclidian Equal Area Gunawan et al. (2021) Topologically (equal area) Work Based (MV & DD) Euclidian Equal Area This Paper Topologically (equal area) Work Based (MV & DD), Duration Based (TD) Manhattan Unequal Area Notations: MV: Volume of materials that have to be transported, DD: Distance between departments, TD: Duration of material transport process Algorithm 1 Pseudocode from NSGA-Ⅱ START Desired number of generations = n_total Generation = n Starting with n = 1 Generating an initial population (Pn) of x individuals Conduct evaluations on the initial population As long as n ≤ n_total: Creating an offspring population of x individuals through the process of crossover and mutation Combining the initial population (Pn) and the derived population (Qn) into a combined population (Rn) Perform ranking (nondominated sorting) on Rn into several fronts (F) Take x/2 top individuals in Rn to become the next population (Pn+1): If the (x/2+1)th and (x/2−1)th individuals are not in the same rank: Continue If the (x/2+1)th and (x/2−1)th individuals are in the same rank (Fz): Perform crowding distance procedures on Fz and sort them Selecting some individuals at Fz amounts to a loss of Pn+1 Pn+1 has been fulfilled n plus 1 FINISH Table 7 Characteristics of best solutions in the last iteration
Sequence of departments "Work Based" MHC "Duration Based" MHC [25, 24, 23, 1, 16, 22, 9, 2, 6, 21, 20, 8, 4, 3, 7, 19, 12, 11, 15, 10, 18, 5, 14, 13, 17] 163 670 WBM-unit 34 750 DBM-unit -
Allahyari MZ, Azab A (2018) Mathematical modeling and multi-start search simulated annealing for unequal-area facility layout problem. Expert Syst. Appl. 91: 46–62. https://doi.org/10.1016/j.eswa.2017.07.049 Besbes M, Zolghadri M, Affonso RC (2021) A method to solve 2D Facility Layout Problem with equipment inputs/outputs constraints using meta-heuristics algorithms. Procedia CIRP 104: 1698–1703. https://doi.org/10.1016/j.procir.2021.11.286 Besbes M, Zolghadri M, Costa Affonso R, Masmoudi F, Haddar M (2020) A methodology for solving facility layout problem considering barriers: genetic algorithm coupled with A* search. J. Intell. Manuf. 31: 615–640. https://doi.org/10.1007/s10845-019-01468-x Braglia M, Zanoni S, Zavanella L (2003) Layout design in dynamic environments: Strategies and quantitative indices. Int. J. Prod. Res. 41: 995–1016. https://doi.org/10.1080/00207540210162983 Choi M, Kim SH, Chung H (2017) Optimal shipyard facility layout planning based on a genetic algorithm and stochastic growth algorithm. Ships Offshore Struct. 12: 486–494. https://doi.org/10.1080/17445302.2016.1176294 Deb K, Pratap A, Agarwal S, Meyarivan T (2002) A fast and elitist multiobjective genetic algorithm: NSGA-Ⅱ. IEEE Trans. Evol. Comput. 6: 182–197. https://doi.org/10.1109/4235.996017 Deep K (2020) Facility layout design using genetic algorithm approach. IUP J. Oper. Manag Gonçalves J, Resende M (2015) A biased random-key genetic algorithm for the unequal area facility layout problem. Eur. J. Oper. Res. 246. https://doi.org/10.1016/j.ejor.2015.04.029 Guan C, Zhang Z, Liu S, Gong J (2019) Multi-objective particle swarm optimization for multi-workshop facility layout problem. J. Manuf. Syst. 53: 32–48. https://doi.org/10.1016/j.jmsy.2019.09.004 Gunawan, Utomo ASA, Benediktus HSV (2021) Optimization of shipyard layout with material handling cost as the main parameter using genetic algorithm. AIP Conf. Proc. 2376. https://doi.org/10.1063/5.0063890 Hani Y, Amodeo L, Yalaoui F, Chen H (2007) Ant colony optimization for solving an industrial layout problem. Eur. J. Oper. Res. 183: 633–642. https://doi.org/10.1016/j.ejor.2006.10.032 Junior WA, Azzolini FGP (2019) Evolutionary algorithm for optimization regarding the planning of topological facilities in layout of a shipyard. IEEE Lat. Am. Trans. 17: 1491–1500 https://doi.org/10.1109/TLA.2019.8931143 Junior WA, Azzolini FGP, Mundim LR, Porto AJV, Amani HJS (2023) Shipyard facility layout optimization through the implementation of a sequential structure of algorithms. Heliyon 9, e16714. https://doi.org/10.1016/j.heliyon.2023.e16714 Kochhar JS, Foster BT, Heragu SS (1998) HOPE: A genetic algorithm for the unequal area facility layout problem. Comput. Oper. Res. 25: 583–594. https://doi.org/10.1016/S0305-0548(97)00100-7 Komarudin, Wong KY (2010) Applying ant system for solving unequal area facility layout problems. Eur. J. Oper. Res. 202: 730–746. https://doi.org/10.1016/j.ejor.2009.06.016 Kulturel-Konak S, Konak A (2011) Unequal area flexible bay facility layout using ant colony optimisation. Int. J. Prod. Res. 49: 1877–1902. https://doi.org/10.1080/00207541003614371 Liu J, Zhang H, He K, Jiang S (2018) Multi-objective particle swarm optimization algorithm based on objective space division for the unequal-area facility layout problem. Expert Syst. Appl. 102: 179–192. https://doi.org/10.1016/j.eswa.2018.02.035 Liu J, Liu J (2019) Applying multi-objective ant colony optimization algorithm for solving the unequal area facility layout problems. Appl. Soft Comput. J. 74: 167–189. https://doi.org/10.1016/j.asoc.2018.10.012 Mak KL, Wong YS, Chan FTS (1998) A genetic algorithm for facility layout problems. Comput. Integr. Manuf. Syst. 11: 113–127. https://doi.org/10.1016/S0951-5240(98)00018-4 Matson JO, Mellichamph JM, Swaminathan SR (1992) EXCITE: Expert consultant for in-plant transportation equipment. Int. J. Prod. Res. 30: 1969–1983. https://doi.org/10.1080/00207549208948133 Muhayat H, Utamima A (2024) Solving unequal area facility layout problems with differential evolution and particle swarm optimization. Procedia Comput. Sci. 234: 302–309. https://doi.org/10.1016/j.procs.2024.03.004 Palomo-Romero JM, Salas-Morera L, García-Hernández L (2017) An island model genetic algorithm for unequal area facility layout problems. Expert Syst. Appl. 68: 151–162. https://doi.org/10.1016/j.eswa.2016.10.004 Palubeckis G (2017) Single row facility layout using multi-start simulated annealing. Comput. Ind. Eng. 103: 1–16. https://doi.org/10.1016/j.cie.2016.09.026 Rosenblatt MJ (2013) Material handling BT-encyclopedia of operations research and management science, in: Gass, S. I., Fu, M. C. (Eds. ). Springer US, Boston, MA, 945–948. https://doi.org/10.1007/978-1-4419-1153-7_590 Samarghandi H, Taabayan P, Jahantigh FF (2010) A particle swarm optimization for the single row facility layout problem. Comput. Ind. Eng. 58: 529–534. https://doi.org/10.1016/j.cie.2009.11.015 Sulaiman SS, Leela Jancy P, Muthiah A, Janakiraman V, Gnanaraj SJP (2021) An evolutionary optimal green layout design for a production facility by simulated annealing algorithm. Mater. Today Proc. 47: 4423–4430. https://doi.org/10.1016/j.matpr.2021.05.256 Turgay S (2018) Multi objective simulated annealing approach for facility layout design. Int. J. Math. Eng. Manag. Sci. 3: 365–380. https://doi.org/10.33889/ijmems.2018.3.4-026 Türk A, Gürgen S, Ozkok M, Altin İ (2022) A comprehensive investigation into the performance of genetic algorithm for effective shipyard topological layout. Proc. Inst. Mech. Eng. Part M J. Eng. Marit. Environ. 236: 726–740 Wang MJ, Hu MH, Ku MY (2005) A solution to the unequal area facilities layout problem by genetic algorithm. Comput. Ind. 56: 207–220. https://doi.org/10.1016/j.compind.2004.06.003 Wong KY, Komarudin (2010) Solving facility layout problems using flexible bay structure representation and ant system algorithm. Expert Syst. Appl. 37: 5523–5527. https://doi.org/10.1016/j.eswa.2009.12.080 Yu-Hsin Chen G (2013) A new data structure of solution representation in hybrid ant colony optimization for large dynamic facility layout problems. Int. J. Prod. Econ. 142: 362–371. https://doi.org/10.1016/j.ijpe.2012.12.012 Zouein PP, Kattan S (2022) An improved construction approach using ant colony optimization for solving the dynamic facility layout problem. J. Oper. Res. Soc. 73: 1517–1531. https://doi.org/10.1080/01605682.2021.1920345