Shuttle-Based Storage and Retrieval System: A Literature Review

Abstract

While the e-commerce logistics industry is developing rapidly, its sustainable development has received certain attention. The ultimate goal of sustainable development is to achieve common, coordinated, fair, efficient, and multi-dimensional development. To promote the sustainable development of logistics, the automation technology of warehousing is undoubtedly an excellent breakthrough, since the automation technology can not only make the warehousing system efficient and with a low-error rate, but also affect the energy consumption of the warehousing system. This paper studies the Shuttle-Based Storage and Retrieval System (SBS/RS) in automated warehousing. Moreover, the paper classifies the existing literature into three categories: Physical design (including depth, configuration, and number of tiers); control strategy (including scheduling rules, storage strategies, scheduling command, and interference); and performance evaluation (including throughput, time, and energy). These categories are all factors that warehouse designers must consider when designing a system. Finally, this paper proposes future research directions for SBS/RS: Adding speed metrics, limiting buffer size, targeting space constraints, applying four-way shuttles, and increasing the carrying capacity of a single machine.

1. Introduction

In recent years, the e-commerce logistics industry has started to attach importance to sustainable development. The research purpose of sustainable development is to seek a balanced development path between economic interests, social harmony, and environmental protection. Efficient logistics strategies can help enterprises improve market competitiveness, save costs, save energy, and reduce emissions [1]. However, there is a large number of manual work in the current warehousing system. This situation in the warehousing system will lead to high error rates and low efficiency. A simple warehousing system failure may interrupt important work, which will lead to high cost [2]. The automatic storage system is more precise, not only can it reduce costs by avoiding errors, but it also improves efficiency. Therefore, automation technology is very important to the sustainable development of the storage system.

At present, the storage and picking technology of warehouses is mainly in the form of an Automated Storage and Retrieval System (AS/RS), which can improve space utilization and productivity. Several systems have been developed through AS/RS: Autonomous Vehicle Storage and Retrieval Systems (AVS/RS), SBS/RS, Vertical Lift Modules (VLM), etc. AVS/RS is a system based on autonomous vehicles [3], which, compared with traditional AS/RS, allows users to flexibly change the number of vehicles in the system to obtain the required throughput [4]. SBS/RS is a shuttle-based system, which, compared with AS/RS, supports high throughput capacity. Finally, VLM is a pallet-based distribution system, which has high operating speed as well as vertical high density storage capacity [5].

Among the literature reviews on automated warehousing [5,6,7,8,9], only a few have mentioned SBS/RS. Among these is Azadeh [10], who described SBS/RS as classified into Aisle-Based Shuttle Systems and Grid-Based Shuttle Systems. The literature on Aisle-Based Shuttle Systems is classified into the following categories: Single/double-Deep Storage, system analysis, design optimization, operations planning and control. The literature on Grid-Based Shuttle Systems is classified into the following branches: System analysis, design optimization, operations planning and control.

This paper mainly reviews and studies SBS/RS and includes some related systems. It has three key differences from previous literature reviews. First, this paper mainly describes SBS/RS, while other review articles do not describe the system in detail. Second, the scope of the literature reviewed in this paper focuses on the years after 2018, while the previous literature involving SBS/RS is based on the literature before 2018. Third, this paper is the first to classify SBS/RS into the following three categories according to the factors considered in system design: Physical design, control strategy, and performance evaluation. Fourth, this paper summarizes the impact of different categories in SBS/RS on sustainability. The specific contents are shown in Table A1.

2. Methodology

This paper takes the English journal literature of the past 5 years (2018–2022) in the Social Sciences Citation Index database in the Web of Science database as the research sample. First, the literature was searched by subject and abstract, and the document type was limited to papers. Second, keywords, such as shuttle warehouse, SBS/RS, and AS/RS, were used. As of October 2022, 238 articles were retrieved. Third, after manual screening of 238 literatures (see Table 1 for screening criteria), 27 papers were obtained. Then, the 27 papers are preliminarily classified, and the framework is sorted out (see Section 3). Finally, according to the keywords in the framework, an extended search was conducted in Connected Paper (http://connectedpapers.com, accessed on 9 October 2022) as well as Google Academic (http://scholar.google.com, accessed on 9 October 2022), and both received seven articles. After the above operation, 41 relevant papers were finally obtained for analysis. The whole process is shown in Figure 1.

The relationship between the publication year and the literature number is shown in Figure 2.

3. Results

3.1. Concept of SBS/RS

With the development of e-commerce, current logistics orders have a demand for a large variety of products, small quantities, and short response times [11]. To better meet these unique order requirements, SBS/RS has started to be applied. SBS/RS is a shuttle-based automatic system [11], whose basic components are Storage Racks (SR), elevators, shuttle carriers, buffer locations, Input/Output (I/O) locations, and Accumulation Conveyors (AC) [6]. In fact, the shuttle can be regarded as an intelligent handling robot [12]. The SR has multiple storage tiers, and the horizontal movement and access of goods can be completed by shuttle vehicles [13]. The emergence of this system with a shuttle can make automated warehousing faster and more convenient. In this paper, the literature on SBS/RS is classified into three categories: Physical design, control strategy, and performance evaluation. The physical design includes the depth, configuration, and number of tiers. Control strategies include scheduling rules, storage strategies, scheduling command, and interference. Performance evaluations include throughput, time, and energy. The overall framework is shown in Figure 3.

3.2. Physical Design

3.2.1. Depth

The depth refers to the width of SR. Generally, single-deep or double-deep SR are used in the system. Double-deep SR are generally designed to be stored side-by-side in two rows. Compared with single-deep SR, double-deep SR save a 50% aisle space [6], allowing them to store more products in the same warehouse area. The application of a single-deep or double-deep system depends on the required warehouse capacity, throughput, etc. If the warehouse is in high demand and the SR layout is relatively small, it is recommended to consider a double-deep system.

Lerher [6] proposed a travel time model to calculate the cycle times of double-deep SBS/RS. The model can calculate the cycle times of Single-Command Cycle (SCC) and Double-Command Cycle (DCC), in order to evaluate the performance of double-deep SBS/RS through cycle times. The results show that the double-deep system does have better performance in some cases. Eder [14] proposed a multi-deep and multi-tier SBS/RS. The system saves floor space due to multi-deep storage. This leads to more complicated transportation of SBS/RS, possibly due to the high load on the shuttle, which in turn can lead to higher costs. To reduce the number of shuttles to control costs, a computational method is needed to describe this system efficiently. This paper evaluates the throughput and other performance of SBS/RS through a continuous-time, Open-Queuing-Network (OQN) system based on limited capacity. In AVS/RS, similar to SBS/RS, some scholars have proposed a multi-deep system [15]. This article provides an analytical model for Depth-First (DF) storage and DF relocation strategy, which allows storage system designers to evaluate rack performance of multi-deep AVS/RS without the need for simulation results. However, since the DF storage strategy is to place the goods in the deepest empty position, this is in reality not an optimal choice. In another paper presented by Marolt [16], a new simulation model is proposed based on three storage methods: Nearest-neighbor storage method, DF method, and random storage method. In multi-deep AVS/RS, nine strategies can be combined. After calculating the cycle time of the multi-deep system under these nine strategies and DCC, this model can help the storage system designer to quickly analyze the performance of the storage system to determine the most suitable strategy combination for the storage system. The literature content of depth is shown in Table 2.

3.2.2. Configurations

Different configurations can lead to different costs; thus, to reduce costs to the greatest extent possible, the physical configuration needs to be determined at an early stage. The configuration of the shuttle system usually includes the location and number of I/O points, the location and number of elevators, and the location and number of shuttles.

Lerher [17] proposed an analytical model. This analytical model has good performance in designing aisle changing shuttle carriers and allows warehouse designers to make decisions at an early stage of SBS/RS establishment. As mentioned by Chen [18], in the case of AS/RS, the different positions of the I/O points will lead to different cycle times. In 3D compact AS/RS, Xu [19] compared the cycle times in the SCC vs. DCC case on the basis of lower I/O point. The results show that using DCC rather than SCC can reduce the cycle times of a single command by about 20%. In subsequent research, Xu [20] proposed four I/O point strategies: Input and output points at two sides, input and output points elevated at one side, input and output points elevated at two sides, and input and output points elevated at a midpoint. After analyzing four different strategies by building a travel time model, Xu found that the I/O point strategy with input and output points elevated at a midpoint performed best in terms of expected travel time. In addition to the I/O points that will cause the final cycle times to vary, the elevator has a certain impact. For example, to maximize throughput and minimize cycle times, Zhao [21] proposed 81 rack schemes based on multiple elevators and determined the optimal scheme through simulation. Regarding the number of shuttles, Ha and Chae [22] explored a decision model for determining the number of shuttles on top of tier-to-tier SBS/RS. This decision model is based on the cycle times of the shuttle and elevator, and it is validated by comparing the results of the analytical and simulation models. This decision model can help warehouse system designers accurately adjust the number of shuttle round trips based on specific throughput requirements. The literature content of configurations is shown in Table 3.

3.2.3. Number of Tiers

In previous studies, single-tier shuttles are usually the mainstay, but setting up shuttles on each floor will greatly increase the cost. To reduce costs, the concept of double-tier or multi-tier shuttles has been proposed. In a single-tier shuttle warehouse, there is a single shuttle for all tiers. In a double-tier or multi-tier shuttle warehouse, one shuttle can complete two or more tiers of commands.

Tappia [23] has studied the single-tier and multi-tier shuttles, respectively. For a single-tier system, numerical results show that the expected operating cycle time to shuttle depth ratio is approximately 1.25, independent of the number and arrival rate of shuttles. Conversely, in a system with multi-tier shuttles, there is no trade-off between the expected operating cycle time and the depth of the shuttle. Therefore, the multi-tier shuttle has certain advantages. There are many studies based on multi-tier shuttles. For example, Dong [24] studied different storage strategies and shuttle scheduling rules of tier-to-tier SBS/RS in SCC or DCC. The mathematical model of the traditional single-tier shuttle storage system does not match the mathematical model of the multi-tier shuttle storage system, since the characteristics of multi-tier parallel retrieval and progressive vertical movement destroy the foundation of the single-tier shuttle storage system. To solve this problem, Wang [25] proposed a two-stage OQN model that treats shuttles and elevators as servers in different phases, and each aspect of the system performance is analyzed from the Shuttle Waiting Period (SWP) and Lift Idle Period (LIP) in the cycle times. The research results show that the model can effectively reduce the total SWP and LIP, which can effectively improve the utilization rate of all equipment, thereby improving the throughput of the entire storage system. To minimize the cycle times of the multi-tier shuttle system, Wang [26] established a mixed integer programming model for the retrieval task scheduling problem for the Multi-tier Shuttle Warehousing System (MSWS). This hybrid model can help provide a reliable reference for MSWS design decisions. The literature content of tier is shown in Table 4.

In a sustainable physical design, we consider that the double-deep saves a 50% aisle space. It not only reduces the cost, but also uses less concrete in warehouses due to less floor space, which can effectively reduce carbon dioxide emissions [27]. Double-tiers or multi-tiers can reduce shuttle. Without the operation of these machines, the energy consumption of the whole system may be reduced.

3.3. Control Strategy

3.3.1. Scheduling Rules

Scheduling rules refer to assigning tasks to shuttles with certain rules. Scheduling of shuttles and elevators becomes a concern to improve efficiency as well as maximize throughput and minimize the cycle times, as different command assignments lead to different efficiencies. The general system will follow the First-In-First-Out (FIFO) rule. This rule operates according to the chronological order of items entering the shelf and according to the principle of first-in-stock items, which is first out of the warehouse when leaving the warehouse.

For example, Ekren [11] proposed an analytical model for FIFO-based scheduling rules to estimate the mean and variance of elevator and shuttle running times, thereby evaluating the performance of storage systems. Wang [28] proposed a time sequence mathematical model according to the movement characteristics of shuttles and elevators. Through the analysis of the mathematical model, the task scheduling problem between the shuttle and the elevator is transformed into an assembly line parallel job problem to generate the scheduling task queue model within the specified time window. At the same time, the Pareto optimal solution based on an elitist non-dominated sorting genetic algorithm is also used to solve the multi-objective optimization function in the task scheduling problem. The final research results show that the optimized scheduling scheme can reduce the SWP and LIP. Yang [29] studied the scheduling problem of goods location allocation and access in MSWS. To maximize the operating efficiency of the system, this paper establishes two related models (integer programming model and dynamic programming model) to solve this problem. In a similar AS/RS, Yang [30] utilized a hybrid polychromatic sets theory-based Genetic Algorithm (GA) to solve the scheduling problem of storage and retrieval. Through experiments, the research results show that this hybrid genetic algorithm has clear advantages in solving the scheduling problem of storage and retrieval. This hybrid algorithm not only reduces the cycle times, but also greatly reduces the computational time required to determine the running route. Baardman [31] studied the scheduling problem of an end-of-aisle automatic storage and retrieval system. The researchers equivalent this problem as a Multiple Traveling Salesman Problem for visiting the same number of cities (mTSPEV); this problem is a Multiple Traveling Salesman Problem (mTSP). At the same time, this paper also proposes a branch-and-cut method and several heuristic algorithms. Finally, the paper proves that this method has good performance in solving scheduling problems. The literature content of scheduling rules is shown in Table 5.

3.3.2. Storage Strategies

When the goods are stored in a reasonable location, the efficiency of goods in and out of the warehouse can be improved, and the cost of warehouse management can be reduced. Therefore, warehouse storage of goods requires a strategy. Storage strategies are generally classified into random storage strategies, classified storage strategies, and nearest-neighbor storage strategies. In this article, we mainly discuss the random storage strategy and classified storage strategy. The random storage strategy indicates that the location where each good is assigned to be stored is not fixed, and it follows a principle that is often changed and randomly generated. The biggest advantage of this strategy is that, since the storage locations can be shared, it only needs to be designed according to the maximum stock quantity of all goods in stock, and the storage area space is thus more efficiently used. However, this approach also creates some drawbacks. For example, due to the random storage, the commonality of the goods is not considered, thus it is very likely that the goods with higher demand and turnover rate are stored at a location far from the output point. The occurrence of this situation has led to an increase in the handling distance in and out of the warehouse.

The classified storage strategy indicates that the warehousing system has already classified each area differently according to certain characteristics of the goods before storage. Then, when the goods are stored, it is necessary to classify the goods according to certain characteristics of the goods and then store the goods in the area corresponding to these characteristics. In the classification area of goods with different characteristics, the allocation of cargo spaces is random. There are some advantages of this classification of storage. First, under this strategy, it is very convenient to access and store the best-selling goods, which can save a great deal of total running time. Second, since each storage area is classified according to the characteristics of the goods, the management of the goods is very convenient, which undoubtedly reduces the rate of cargo damage and the probability of major dangers. However, at the same time, classified warehousing also has certain disadvantages. The classified storage strategy needs to estimate the maximum quantity of each type of good and design the maximum stock quantity accordingly. However, in actual situations, it is very likely that the goods cannot be placed in the warehouse, thus this situation leads to a low utilization rate of the warehouse. Many scholars have researched these two storage strategies.

Dong [24] established the travel time model under different combinations of storage rules based on the tier-to-tier shuttle system to calculate the cycle times. The comparison of the results leads to the following conclusions. Classification storage is better than random storage; however, when the demand information of warehouse designers or managers is limited, resulting in information asymmetry or when the demand rate of goods storage volume fluctuates greatly, the stochastic storage strategy is more suitable. In the case of information asymmetry, the designer or manager cannot understand the information about the goods to be stored and does not know its characteristics and demand rate, thus it is impossible to carry out a targeted partition of the warehouse according to the characteristics of the goods. Therefore, although random storage will lead to lower space utilization, it still has certain advantages in this case. Eder [32] proposed that one way to achieve higher throughput for SBS/RS is to change the storage strategy to a classified storage strategy. This study models the system as a continuous-time OQN system with limited capacity and presents an example of how to use this computational model to optimize the existing SBS/RS by classified storage strategy. Moreover, in this example, depending on the configuration of the policy, applying the classified storage to the existing SBS/RS more than doubled the throughput of a system that did not use the classified storage strategy. In AS/RS, Yang [33] studied a joint optimization in a multi-shuttle system under a shared storage strategy (multiple cargoes stored in a storage stack). Overall, different systems have different preferences for storage strategies under different configuration strategies (SCC or DCC, single-deep or double-deep, etc.). The system designer cannot choose the storage strategy casually when designing the system and should make a careful choice according to other configurations of the system. The literature content of storage strategies is shown in Table 6.

3.3.3. Scheduling Commands

Scheduling commands are classified into SCC and DCC. SCC indicates that the shuttle of SBS/RS only performs one of the commands between the storage and retrieval in one cycle. DCC refers to two commands in which the shuttle of SBS/RS performs storage and retrieval at the same time in one cycle. Wu [34] constructed an OQN model for SBS/RS based on DCC and used this model to find the minimum facility cost configuration of SBS/RS given the storage capacity, throughput, and order cycle time requirements. In addition, this article compares SBS/RS and robotic systems in terms of facility cost, construction cost, and operating cycle time, which can help warehouse designers make sound decisions when choosing a system.

Studying similar AS/RS, Xu [35] proposed a double-deep and multi-aisle automatic storage system. The system is based on a random storage strategy and rearrangement of blocking loads. In this case, the paper establishes the SCC and DCC travel time models under three rearrangement rules. To validate the model, Xu performed simulated and numerical experiments. The research results show that with the increase in the number of service channels of the shuttle, the cycle times of the shuttle show an increasing trend, and DCC is better than SCC. Regarding the similar AVS/RS system, Manzini [3] studied a multi-tier AVS/RS system. In order to be used to determine the cycle times and operating distance of deep-lane unit-loads and multi-tier AVS/RS systems, an SCC- and DCC-based travel time model under different configurations is proposed. To verify the validity of the model, the research takes an actual warehouse as an example to conduct an example analysis, and the validity of the model is verified by simulation. Moreover, it presents useful guidelines for system layout configuration. Furthermore, Lerher [36] proposed a travel time model based on SCC and DCC that can be used to calculate the cycle times and throughput performance of multi-tier AVS/RS. On the contrary, Lerher builds a simulation model that can be used to evaluate the accuracy of the proposed travel time model. However, in addition to these two scheduling commands, there is actually an improved 2n-command [37]. Under this command cycle, Yang discusses the integrated optimization problem of the shuttle’s position allocation and sequencing in AS/RS. To solve this problem, an integer quadratic programming model is established to describe this ensemble optimization problem. For small problems, the model can be solved by CPLEX to obtain the optimal cycle times of multi-shuttle vehicles. For large problems, two tabu search algorithms are proposed. First-come, first-serve and nearest-neighbor algorithms are used to generate initial solutions. Most importantly, this study analyzes the experimental results from a practical application perspective and presents a list of parameters for applying the proposed heuristic algorithm in different practical scenarios. The literature content of scheduling commands is shown in Table 7.

3.3.4. Interference

Mutual interference here refers to when two elevators share a mast, making it impossible for the elevators to pass each other [38]. In this case, if there is no reasonable scheduling rule, the two elevators will collide if the upper elevator receives a downward command at the same time that the lower elevator receives an upward command. Once the two elevators collide, the system will temporarily stop working, placing a significant burden on the warehouse. Zhao [39] proposes that if the problem of elevators interfering with each other can be solved, the storage system will be able to achieve higher throughput. Therefore, by considering the collision-free trajectory prediction method of acceleration and deceleration, Zhao used the proposed genetic algorithm to sort and allocate elevator requests. This method, referred to as GA + EBTP + A for short, effectively solves the elevator scheduling problem. In a similar AS/RS, Foumani [40] proposes the Traveling Salesman Problem (TSP) to minimize the cycle times, but this method leads to the collision event of the robot, which is similar to two elevators interfering with each other. In this paper, the Cross-Entropy (CE) method is used in this problem stage. To evaluate the performance of the CE method, a computational analysis was performed on various test problems. The results obtained by this method are compared with the optimal solutions obtained by the CPLEX method. The experimental results show that this method can solve the sequencing problem of robots. The literature content of interference is shown in Table 8.

Operation efficiency and operation cost have a crucial impact on sustainable development [41], and the control strategy is one of the factors that affect operation efficiency and operation cost. We figure that reasonable scheduling rules can avoid the shuttle from taking redundant routes. This not only reduces the energy consumption caused by the shuttle movement, but also improves the operation efficiency, thus improving the sustainability. For the storage strategy, when allocating items to locations with the minimum movement priority in the system, the energy cost is minimized, thereby helping energy sustainability [42]. For mutual interference, we consider that since the elevators will not collide, the waste of resources will be avoided to improve the sustainability.

3.4. Performance Evaluations

An SBS/RS can be evaluated from a number of perspectives, the most common of which are throughput, time, and energy consumption.

3.4.1. Throughput

Throughput refers to the total amount of goods stored and retrieved by a warehouse over a certain period of time. In most studies, throughput is used as an important indicator to evaluate system performance. Generally, the larger the throughput, the better, thus most scholars conduct research to improve the throughput of the system. Tappia [43] used a Semi-Open Queuing Network (SOQN) model in SBS/RS to study the impact of a novel analytical model for integrated storage and order picking systems. Azadeh [44] has carried out some research on a vertical automated warehouse (i.e., where the shuttle can move vertically without elevators). On the one hand, a Closed Queuing Network (CQN) model is established without considering shuttle congestion and Approximate Mean Value Analysis (AMVA) is used to estimate the system throughput. On the other hand, when considering the blocking delay, the study adopts two strategies for coping with shuttle blocking: Robot Recirculation (REC) and Wait-on-Spot (WOS). In the case of these two strategies, the throughput of the system is calculated through the model. Finally, comparing the results of the two strategies, the WOS strategy has a slight advantage when the number of shuttles in the warehousing system is small. When the number of shuttles increases, the system throughput will drop sharply; in this case, adopting the REC strategy can alleviate this problem. A similar problem was studied by Chen [45]. However, in this paper, the researchers use a SOQN to study this problem. The final conclusions are similar to those reached by Azadeh. Yang [46] modeled and evaluated another Robotic Mobile Fulfilment System (RMFS). In this paper, a SOQN model is built and solved using AMVA. By solving the model, approximate analytical solutions for system throughput, shuttle utilization, and queue length can be obtained. Subsequently, the model was verified and evaluated through simulation. The literature content of throughput is shown in Table 9.

3.4.2. Time

Cycle time refers to the time it takes for the warehouse to perform a storage or retrieval. Cycle time is also an important indicator for evaluating system performance.

Ekren [13] proposed a graph-based solution. The solution involves collecting a large number of graphs of different design scenarios to visualize the performance values of the studied SBS/RS. On the one hand, since the provided diagram contains thousands of SBS/RS design solutions, the solution allows warehouse designers or managers to define and evaluate system performance in a timely manner. On the other hand, these design schemes not only include SR design, aisles, tiers, etc., but also are evaluated based on the average utilization rate of the elevator and the average cycle times of the system, giving them a certain degree of accuracy. Foumani [40] applied the TSP to a similar AS/RS. The decision to be made in this problem involves finding the optimal order sequence and the optimal sequence of items in each order. These decisions are made to minimize the total cycle time. Antonio [47] evaluates the performance of deep lane AVS/RS by considering three criteria, one of which is the cycle time. An analytical model is proposed to evaluate the performance of the system, and then different warehouse designs are simulated and verified under different scenarios. Ultimately, of course, the purpose of these models is to allow warehouse designers to evaluate the performance of their systems considering multiple real-world scenarios. Yener [48] utilizes data mining techniques to design warehouses to determine the effectiveness of shipping distance and average cycle times. Moreover, an integer linear mathematical model based on the Vehicle Routing Problem (VRP) is used to route a large number of randomly selected picking requests. In a subsequent instance, the results of the study showed that the total wait time for orders in the system decreased by 2.5 min after using VRP. Nicolas [49] focuses on another VLM, studying the problem of batch processing of orders in VLM. To minimize the cycle times, the paper provides an order batch optimization model and solves it with CPLEX. To test and validate the accuracy of the model, real data from two companies operating in different industries are used. Numerical experiments show that this model has certain advantages over the performance of batch processing methods currently used by these companies. Moreover, CPLEX is only suitable for dealing with small sample sizes. For complex cases that cannot be solved with CPLEX, the article develops a metaheuristic approach that usually produces very good solutions in less than a minute. The literature content of time is shown in Table 10.

3.4.3. Energy

Energy consumption refers to the consumption and regeneration of energy in a storage or retrieval transaction. For the sustainable development of logistics, researchers should pay attention to energy efficiency. If the efficiency of energy use is increased, the cost of energy use will be reduced, and the release of harmful gases, such as carbon dioxide can be reduced. Therefore, it is important to take energy consumption into consideration in the SBS/RS studied herein.

Meneghetti [50] concluded that if one wants to reduce the energy consumption of the system, certain changes can be made in the shape of the SR. Since the early AS/RS, elevators have not been widely used. In this early system, the stacking crane acts as an elevator. Since the specification of the energy consumption of the stacking crane is strictly related to the height of the SR, we can calculate the optimal SR shape through the model to reduce the energy consumption required by the crane. In this study, a simulation analysis was performed to evaluate the impact of SR shape on energy saving. The results show that, regardless of the demand curve and optimization goals, intermediate height SR shapes perform best in terms of energy efficiency, while lower SR shapes perform better when the cycle times performance is considered. Hahn-Woernle [51] proposed a Power-Load Management (PLM) method that avoids power peaks, as uncontrolled power peaks in automated warehouses lead to higher costs. The article assumes that power peaks can be significantly reduced by delaying tasks without impacting throughput. The final study results show that power consumption peaks can be avoided with little throughput loss under a maximum power limit and a mean power limit. Therefore, this method of PLM is an effective way to reduce energy peaks in automated warehouses. Some scholars also control the energy consumption by changing the design parameters of the system. Moreover, Ekren [11] proposed an analytical model tool that can calculate the average energy consumption and energy regeneration for each storage or retrieval in the system. This tool allows for changing input design parameters (e.g., discrete travel lengths, velocity of vehicles, etc.) to evaluate system performance, thereby controlling the consumption of energy. In the work of Ekren [12], another tool was proposed based on an OQN model. This tool can also estimate some important performance metrics in SBS/RS design. Therefore, controlling energy consumption can reduce the cost of the storage system, in order that energy consumption can be taken into account when designing the system. The literature content of energy is shown in Table 11.

The energy consumption is very important for the sustainable development of warehouses. We believe that energy consumption can be controlled from the rack shape, power, and other aspects. The energy consumption is reduced in order that the warehouse can achieve the sustainability.

4. Conclusions and Future Research

4.1. Conclusions

Among the drivers for sustainability is automation, which is crucial to the long-term sustainable development goals [52]. There are many factors affecting sustainable development in the current logistics warehousing system. On the one hand, there are low efficiency, high error rate, and high cost, but the application of automation technology can solve these problems very well. In addition, using automation in warehouses is considered as a green solution [52]. On the other hand, automation technology is related to energy consumption, and few articles have made some progress in this field. Therefore, this paper mainly studies the SBS/RS literature in automated warehousing.

This paper classifies the literature on SBS/RS into three categories: Physical design, control strategies, and performance evaluation. First, for physical design, double-deep shuttles and multi-tier shuttles are ideal choices, since this can not only reduce the aisle to save energy consumption, but also reduce the floor area to lower carbon dioxide emissions. The configuration of shuttles needs to be strategized according to the characteristics of the system. Second, for control strategies, classified storage strategies and DCC can make the storage system more efficient, this high efficiency is conducive to sustainable development. Moreover, scheduling rules can be selected according to requirements. Choosing the most appropriate rules can avoid the redundant movement of shuttles, which can reduce energy consumption and promote sustainability. Third, throughput, time, and energy are all important indicators for performance evaluation. Designers can design the sustainable system by setting different importance levels of the three indicators according to their own needs.

However, the article found that these three categories influence each other to some extent. Therefore, if the designer of the warehouse system wants to design the most reasonable sustainable warehouse, he needs to choose different strategies according to the performance requirements of the designed system.

4.2. Future Research

After reviewing the SBS/RS literature, this paper concludes that the future research directions of the system are as follows.

4.2.1. Speed

In the existing literature, most of the shuttles and elevators are assumed to work under a uniform speed; however, in a real storage system, it is difficult for these machines to run at a uniform speed. Not running at a constant speed may cause the machines to block or result in other issues, these situations can prevent the effective use of resources, thus preventing sustainable development. Therefore, future research should consider SBS/RS at different speeds and should add factors, such as acceleration and deceleration.

4.2.2. Buffer

The existing literature generally assumes an infinite buffer, but this is not possible in a real system. Therefore, if researchers want to explore the SBS/RS more precisely, the reality of a limited buffer should be taken into account in future research.

4.2.3. Space Restrictions

With regard to system performance evaluation, the existing literature generally limits the storage system with the goal of maximizing throughput, minimizing running time, and minimizing energy consumption. It would be advisable for future research to add space constraints, since in real system design the size of the warehouse is generally the first factor to be considered.

4.2.4. Four-Way Shuttle

In most studies, the SBS/RS is set up with one shuttle per floor. However, this approach would lead to an increase in cost, thus in reality, there are double-layer or multi-layer shuttles. However, these shuttles can also potentially be operated in four directions; namely, the shuttles can be operated not only horizontally but also vertically. In this case, elevators are no longer required, which can greatly reduce costs. Although a few studies have examined this point, the research in this area is not yet in-depth. Therefore, more detailed research can be carried out on this point in the future.

4.2.5. Load Capacity

In most studies, the system is examined with the shuttle as single-load. However, in reality, overloading is a wiser choice, since multi-load work can improve work efficiency and reduce running time, thus achieve sustainable development. Therefore, future SBS/RS research can examine multi-load shuttles.

Author Contributions

Conceptualization, Y.L. and Z.L.; methodology, Y.L. and Z.L.; validation, Z.L.; formal analysis, Z.L.; investigation, Z.L.; resources, Z.L.; data curation, Z.L.; writing—original draft preparation, Z.L.; writing—review and editing, Y.L.; visualization, Z.L.; supervision, Y.L.; project administration, Y.L.; funding acquisition, Y.L. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Figure 1. Research design.

Figure 2. Year distribution of literature.

Figure 3. Framework of review.

Table 1

Literature screening criteria.

Number	Exclusion Criteria	Discard the Title	Judgment Method
1	Titles unrelated to automation, system, warehouse, etc.	diseases, electrochemical, medicines, refugees, etc.	Read the title, abstract, keywords
2	Titles unrelated to SBS/RS, shuttle warehouse, AS/RS, etc.	flying warehouse delivery, coastal shuttle, tanker, etc.	Read the title, abstract, keywords
3	Titles related to economics, case studies, etc.	economic methodology, case study, etc.	Read the abstract

Table 2

Depth of shuttles.

References	System	Depth
[6]	SBS/RS	Double-deep
[14]	SBS/RS	Multi-deep
[15]	AVS/RS	Multi-deep
[16]	AVS/RS	Multi-deep

Table 3

Configurations of shuttles.

References	Model	Configurations
[17]	Analytical model	SR, Speed
[18]	Analytical model	I/O point
[19]	Analytical model	I/O point
[20]	Analytical model	I/O point
[21]	Simulation model	SR
[22]	Analytical model	Number

Table 4

Tier of shuttles.

References	Model	Number of Tiers	Objective
[23]	SOQN	Double-tier	Number
[24]	Travel time model	Double-tier	Storage policies Scheduling rulesNumber
[25]	Two-stage OQN	Multi-tier	Throughput
[26]	Mixed integer programming model	Multi-tier	Scheduling rules

Table 5

Scheduling rules of shuttles.

References	Model
[11]	Analytical model
[28]	Time sequence mathematical model
[29]	Integer programming modelDynamic programming model
[30]	Hybrid polychromatic sets theory-based genetic algorithm
[31]	MTSPEV

Table 6

Storage strategies of shuttles.

References	Model	Storage Strategies
[24]	Analytical model	Classified storage strategy
[32]	Continuous-time OQN	Classified storage strategy
[33]	Analytical model	Shared storage strategy

Table 7

Scheduling commands of shuttles.

References	Model	Scheduling Commands
[34]	OQN	DCC
[35]	Travel time model	SCC/DCC
[3]	Travel time model	SCC/DCC
[36]	Travel time model	SCC/DCC
[37]	Integer quadratic programming model	2n-command

Table 8

Interference of shuttles.

References	Model
[39]	GA + EBTP + A
[40]	TSPCE

Table 9

Throughput of shuttles.

References	Model
[43]	Analytical modelSOQN
[44]	CQN
[45]	SOQN
[46]	SOQN

Table 10

Time of shuttles.

References	Solution
[13]	Graph-based solution
[40]	TSP
[47]	Analytical model
[48]	Data mining techniques
[49]	Order batch optimization model

Table 11

Energy of shuttles.

References	Solution
[50]	Change shape of the SR
[51]	PLM
[11]	Analytical model
[12]	OQN

Appendix A

Table A1

Summary of literature.

Reference	Model	Scheduling Commands	Research Content	Sustainability
Manzini et al. (2016) [3]	Travel time model	SCC/DCC	Scheduling commandsDepthTier	EfficiencyEnergySpaceCost
Ekren (2021) [4]	Multi-objective optimization solution		TimeEnergy	EnergyEfficiencyTime
Lerher (2016) [6]	Travel time model	SCC/DCC	DepthTime	TimeSpace
Ekren et al. (2018) [11]	Analytical model	SCC	Scheduling rules EnergyTime	EnergyTime
Ekren and Akpunar (2021) [12]	OQN	SCC/DCC	Scheduling rulesEnergyTime	EnergyEfficiencyTime
Yetkin Ekren (2017) [13]	Graph-based solution	DCC	Time	TimeEfficiency
Eder (2020) [14]	Continuous-time OQN	SCC	DepthTime	SpaceTime
Marolt et al. (2022) [15]	Analytical model	SCC/DCC	DepthStorage strategiesThroughput	SpaceEfficiencyEnergy
Marolt et al. (2022) [16]	Simulation model	DCC	DepthStorage strategiesTime	SpaceTimeEfficiency
Lerher (2018) [17]	Analytical model	SCC/DCC	ConfigurationsTime	Time
W. (Amanda) Chen et al. (2022) [18]	Analytical model	SCC	ConfigurationsTime	Time
Xu et al. (2018) [19]	Analytical model	DCC	ConfigurationsTime	Time
Xu et al. (2020) [20]	Analytical model	SCC	ConfigurationsTime	Time
Zhao et al. (2016) [21]	Simulation model	SCC	ConfigurationsTimeThroughput	Time
Ha and Chae (2019) [22]	Analytical model	DCC	ConfigurationsTime	Time
Tappia et al. (2017) [23]	SOQN	SCC	TierTime	EnergyTime
Dong and Jin (2021) [24]	Travel time model	SCC/DCC	TierStorage strategiesTime	TimeEfficiencyEnergy
Wang et al. (2016) [25]	Two-stage OQN	SCC	TierStorage strategiesTime	TimeEfficiencyEnergy
Wang et al. (2020) [26]	Mixed integer programming model	SCC	TierScheduling rulesTime	TimeEfficiencyEnergy
Wang et al. (2015) [28]	Time sequence mathematical model	SCC	TierScheduling rulesTime	TimeEfficiencyEnergy
P. Yang et al. (2015) [29]	Integer programming modelDynamic programming model	SCC/MCC (Multiple)	Scheduling rules	EfficiencyEnergy
Z. Yang et al. (2019) [30]	Hybrid polychromatic sets theory-based genetic algorithm		Scheduling rules	Efficiency Energy
Baardman et al. (2021) [31]	mTSPEV	SCC/DCC	Scheduling rulesTime	EfficiencyTimeEnergy
Eder (2020) [32]	Continuous-time OQN	SCC/DCC	Storage strategiesThroughput	EfficiencyEnergy
P. Yang et al. (2015) [33]	Analytical model	DCC	Storage strategiesScheduling rules	EnergyEfficiency
Wu et al. (2020) [34]	OQN	DCC	Scheduling commandsTimeThroughput	CostTime
Xu et al. (2016) [35]	Travel time model	SCC/DCC	Scheduling commandsTimeDepth	TimeSpace
Lerher et al. (2021) [36]	Travel time model	SCC/DCC	Scheduling commandsTime	Time
P. Yang et al. (2017) [37]	Integer quadratic programming model	2n-command	Scheduling commandsScheduling rulesStorage strategies	TimeEfficiency
Carlo and Vis (2012) [38]	Integrated look-ahead strategy heuristic	SCC	Interference	Efficiency
Zhao et al. (2018) [39]	GA + EBTP + A		Interference	Efficiency
Foumani et al. (2018) [40]	TSPCE		InterferenceTime	EfficiencyTime
Tappia et al. (2019) [43]	Analytical modelSOQN	SCC	ThroughputTime	Time
Azadeh et al. (2019) [44]	CQN	DCC	Throughput	Cost
W. Chen et al. (2022) [45]	SOQN	SCC/DCC	ThroughputTimeStorage strategies	TimeEfficiency
P. Yang et al. (2021) [46]	SOQN		ThroughputTime	TimeEfficiency
D’Antonio et al. (2018) [47]	Analytical model	MCC	TimeConfigurations	EnergyTime
Yener and Yazgan (2019) [48]	Data mining techniques		Time	EfficiencyTime
Nicolas et al. (2018) [49]	Order batch optimization model		Time	TimeEfficiency
Meneghetti et al. (2015) [50]	Change shape of the SR	SCC/DCC	EnergyTime	EnergyTime
Hahn-Woernle and Günthner (2018) [51]	PLM	SCC/DCC	EnergyThroughput	EnergyCost

References

1. Punyim, P.; Karoonsoontawong, A.; Unnikrishnan, A.; Ratanavaraha, V. A Heuristic for the Two-Echelon Multi-Period Multi-Product Location–Inventory Problem with Partial Facility Closing and Reopening. Sustainability; 2022; 14, 10569. [DOI: https://dx.doi.org/10.3390/su141710569]

2. Pejić, V.; Lerher, T.; Jereb, B.; Lisec, A. Lean and Green Paradigms in Logistics: Review of Published Research. PROMET; 2016; 28, pp. 593-603. [DOI: https://dx.doi.org/10.7307/ptt.v28i6.2078]

3. Manzini, R.; Accorsi, R.; Baruffaldi, G.; Cennerazzo, T.; Gamberi, M. Travel Time Models for Deep-Lane Unit-Load Autonomous Vehicle Storage and Retrieval System (AVS/RS). Int. J. Prod. Res.; 2016; 54, pp. 4286-4304. [DOI: https://dx.doi.org/10.1080/00207543.2016.1144241]

4. Yetkin Ekren, B. A Multi-Objective Optimisation Study for the Design of an AVS/RS Warehouse. Int. J. Prod. Res.; 2021; 59, pp. 1107-1126. [DOI: https://dx.doi.org/10.1080/00207543.2020.1720927]

5. Dukic, G.; Opetuk, T.; Lerher, T. A Throughput Model for a Dual-Tray Vertical Lift Module with a Human Order-Picker. Int. J. Prod. Econ.; 2015; 170, pp. 874-881. [DOI: https://dx.doi.org/10.1016/j.ijpe.2015.04.009]

6. Lerher, T. Travel Time Model for Double-Deep Shuttle-Based Storage and Retrieval Systems. Int. J. Prod. Res.; 2016; 54, pp. 2519-2540. [DOI: https://dx.doi.org/10.1080/00207543.2015.1061717]

7. Jaghbeer, Y.; Hanson, R.; Johansson, M.I. Automated Order Picking Systems and the Links between Design and Performance: A Systematic Literature Review. Int. J. Prod. Res.; 2020; 58, pp. 4489-4505. [DOI: https://dx.doi.org/10.1080/00207543.2020.1788734]

8. Custodio, L.; Machado, R. Flexible Automated Warehouse: A Literature Review and an Innovative Framework. Int. J. Adv. Manuf. Technol.; 2020; 106, pp. 533-558. [DOI: https://dx.doi.org/10.1007/s00170-019-04588-z]

9. Fragapane, G.; de Koster, R.; Sgarbossa, F.; Strandhagen, J.O. Planning and Control of Autonomous Mobile Robots for Intralogistics: Literature Review and Research Agenda. Eur. J. Oper. Res.; 2021; 294, pp. 405-426. [DOI: https://dx.doi.org/10.1016/j.ejor.2021.01.019]

10. Azadeh, K.; De Koster, R.; Roy, D. Robotized and Automated Warehouse Systems: Review and Recent Developments. Transp. Sci.; 2019; 53, pp. 917-945. [DOI: https://dx.doi.org/10.1287/trsc.2018.0873]

11. Ekren, B.Y.; Akpunar, A.; Sari, Z.; Lerher, T. A Tool for Time, Variance and Energy Related Performance Estimations in a Shuttle-Based Storage and Retrieval System. Appl. Math. Model.; 2018; 63, pp. 109-127. [DOI: https://dx.doi.org/10.1016/j.apm.2018.06.037]

12. Ekren, B.Y.; Akpunar, A. An Open Queuing Network-Based Tool for Performance Estimations in a Shuttle-Based Storage and Retrieval System. Appl. Math. Model.; 2021; 89, pp. 1678-1695. [DOI: https://dx.doi.org/10.1016/j.apm.2020.07.055]

13. Yetkin Ekren, B. Graph-Based Solution for Performance Evaluation of Shuttle-Based Storage and Retrieval System. Int. J. Prod. Res.; 2017; 55, pp. 6516-6526. [DOI: https://dx.doi.org/10.1080/00207543.2016.1203076]

14. Eder, M. An Approach for Performance Evaluation of SBS/RS with Shuttle Vehicles Serving Multiple Tiers of Multiple-Deep Storage Rack. Int. J. Adv. Manuf. Technol.; 2020; 110, pp. 3241-3256. [DOI: https://dx.doi.org/10.1007/s00170-020-06033-y]

15. Marolt, J.; Šinko, S.; Lerher, T. Model of a Multiple-Deep Automated Vehicles Storage and Retrieval System Following the Combination of Depth-First Storage and Depth-First Relocation Strategies. Int. J. Prod. Res.; 2022; [DOI: https://dx.doi.org/10.1080/00207543.2022.2087568]

16. Marolt, J.; Kosanić, N.; Lerher, T. Relocation and Storage Assignment Strategy Evaluation in a Multiple-Deep Tier Captive Automated Vehicle Storage and Retrieval System with Undetermined Retrieval Sequence. Int. J. Adv. Manuf. Technol.; 2022; 118, pp. 3403-3420. [DOI: https://dx.doi.org/10.1007/s00170-021-08169-x]

17. Lerher, T. Aisle Changing Shuttle Carriers in Autonomous Vehicle Storage and Retrieval Systems. Int. J. Prod. Res.; 2018; 56, pp. 3859-3879. [DOI: https://dx.doi.org/10.1080/00207543.2018.1467060]

18. Chen, W.; Gong, Y.; de Koster, R.B.M. Performance Estimation of a Passing-Crane Automated Storage and Retrieval System. Int. J. Prod. Res.; 2022; 60, pp. 1210-1230. [DOI: https://dx.doi.org/10.1080/00207543.2020.1854886]

19. Xu, X.; Gong, Y.; Fan, X.; Shen, G.; Zou, B. Travel-Time Model of Dual-Command Cycles in a 3D Compact AS/RS with Lower Mid-Point I/O Dwell Point Policy. Int. J. Prod. Res.; 2018; 56, pp. 1620-1641. [DOI: https://dx.doi.org/10.1080/00207543.2017.1361049]

20. Xu, X.; Zhao, X.; Zou, B.; Gong, Y.; Wang, H. Travel Time Models for a Three-Dimensional Compact AS/RS Considering Different I/O Point Policies. Int. J. Prod. Res.; 2020; 58, pp. 5432-5455. [DOI: https://dx.doi.org/10.1080/00207543.2019.1659519]

21. Zhao, N.; Lei, L.; Saipeng, Z.; Lodewijks, G. An Efficient Simulation Model for Rack Design in Multi-Elevator Shuttle-Based Storage and Retrieval System. Simul. Model. Pract. Theory; 2016; 67, pp. 100-116. [DOI: https://dx.doi.org/10.1016/j.simpat.2016.03.007]

22. Ha, Y.; Chae, J. A Decision Model to Determine the Number of Shuttles in a Tier-to-Tier SBS/RS. Int. J. Prod. Res.; 2019; 57, pp. 963-984. [DOI: https://dx.doi.org/10.1080/00207543.2018.1476787]

23. Tappia, E.; di Milano, P.; Roy, D. Modeling, Analysis, and Design Insights for Shuttle-Based Compact Storage Systems. Trans. Sci.; 2017; 51, pp. 269-295. [DOI: https://dx.doi.org/10.1287/trsc.2016.0699]

24. Dong, W.; Jin, M. Travel Time Models for Tier-to-Tier SBS/RS with Different Storage Assignment Policies and Shuttle Dispatching Rules. Transp. Res. Part E Logist. Transp. Rev.; 2021; 155, 102485. [DOI: https://dx.doi.org/10.1016/j.tre.2021.102485]

25. Wang, Y.; Mou, S.; Wu, Y. Storage Assignment Optimization in a Multi-Tier Shuttle Warehousing System. Chin. J. Mech. Eng.; 2016; 29, pp. 421-429. [DOI: https://dx.doi.org/10.3901/CJME.2015.1221.152]

26. Wang, Y.; Liu, Z.; Huang, K.; Mou, S.; Zhang, R. Model and Solution Approaches for Retrieval Operations in a Multi-Tier Shuttle Warehouse System. Comput. Ind. Eng.; 2020; 141, 106283. [DOI: https://dx.doi.org/10.1016/j.cie.2020.106283]

27. Meneghetti, A.; Monti, L. Sustainable Storage Assignment and Dwell-Point Policies for Automated Storage and Retrieval Systems. Prod. Plan. Control; 2013; 24, pp. 511-520. [DOI: https://dx.doi.org/10.1080/09537287.2011.637525]

28. Wang, Y.; Mou, S.; Wu, Y. Task Scheduling for Multi-Tier Shuttle Warehousing Systems. Int. J. Prod. Res.; 2015; 53, pp. 5884-5895. [DOI: https://dx.doi.org/10.1080/00207543.2015.1012604]

29. Yang, P.; Miao, L.; Xue, Z.; Qin, L. An Integrated Optimization of Location Assignment and Storage/Retrieval Scheduling in Multi-Shuttle Automated Storage/Retrieval Systems. J. Intell. Manuf.; 2015; 26, pp. 1145-1159. [DOI: https://dx.doi.org/10.1007/s10845-013-0846-7]

30. Yang, Z.; Wang, L.; Cai, Y.; Kimie, K. A Polychromatic Sets Theory Based Algorithm for the Input/Output Scheduling Problem in AS/RSs. Evol. Intel.; 2019; 12, pp. 333-340. [DOI: https://dx.doi.org/10.1007/s12065-018-0183-z]

31. Baardman, L.; Roodbergen, K.J.; Carlo, H.J.; Schrotenboer, A.H. A Special Case of the Multiple Traveling Salesmen Problem in End-of-Aisle Picking Systems. Transp. Sci.; 2021; 55, pp. 1151-1169. [DOI: https://dx.doi.org/10.1287/trsc.2021.1075]

32. Eder, M. Analytical Model to Estimate the Performance of Shuttle-Based Storage and Retrieval Systems with Class-Based Storage Policy. Int. J. Adv. Manuf. Technol.; 2020; 107, pp. 2091-2106. [DOI: https://dx.doi.org/10.1007/s00170-020-04990-y]

33. Yang, P.; Miao, L.; Xue, Z.; Ye, B. Variable Neighborhood Search Heuristic for Storage Location Assignment and Storage/Retrieval Scheduling under Shared Storage in Multi-Shuttle Automated Storage/Retrieval Systems. Transp. Res. Part E Logist. Transp. Rev.; 2015; 79, pp. 164-177. [DOI: https://dx.doi.org/10.1016/j.tre.2015.04.009]

34. Wu, Y.; Zhou, C.; Ma, W.; Kong, X.T.R. Modelling and Design for a Shuttle-Based Storage and Retrieval System. Int. J. Prod. Res.; 2020; 58, pp. 4808-4828. [DOI: https://dx.doi.org/10.1080/00207543.2019.1665202]

35. Xu, X.; Zou, B.; Shen, G.; Gong, Y. Travel-Time Models and Fill-Grade Factor Analysis for Double-Deep Multi-Aisle AS/RSs. Int. J. Prod. Res.; 2016; 54, pp. 4126-4144. [DOI: https://dx.doi.org/10.1080/00207543.2015.1106604]

36. Lerher, T.; Ficko, M.; Palčič, I. Throughput Performance Analysis of Automated Vehicle Storage and Retrieval Systems with Multiple-Tier Shuttle Vehicles. Appl. Math. Model.; 2021; 91, pp. 1004-1022. [DOI: https://dx.doi.org/10.1016/j.apm.2020.10.032]

37. Yang, P.; Peng, Y.; Ye, B.; Miao, L. Integrated Optimization of Location Assignment and Sequencing in Multi-Shuttle Automated Storage and Retrieval Systems under Modified 2 n-Command Cycle Pattern. Eng. Optim.; 2017; 49, pp. 1604-1620. [DOI: https://dx.doi.org/10.1080/0305215X.2016.1261128]

38. Carlo, H.J.; Vis, I.F.A. Sequencing Dynamic Storage Systems with Multiple Lifts and Shuttles. Int. J. Prod. Econ.; 2012; 140, pp. 844-853. [DOI: https://dx.doi.org/10.1016/j.ijpe.2012.06.035]

39. Zhao, N.; Luo, L.; Lodewijks, G. Scheduling Two Lifts on a Common Rail Considering Acceleration and Deceleration in a Shuttle Based Storage and Retrieval System. Comput. Ind. Eng.; 2018; 124, pp. 48-57. [DOI: https://dx.doi.org/10.1016/j.cie.2018.07.007]

40. Foumani, M.; Moeini, A.; Haythorpe, M.; Smith-Miles, K. A Cross-Entropy Method for Optimising Robotic Automated Storage and Retrieval Systems. Int. J. Prod. Res.; 2018; 56, pp. 6450-6472. [DOI: https://dx.doi.org/10.1080/00207543.2018.1456692]

41. Cai, J.; Li, X.; Liang, Y.; Ouyang, S. Collaborative Optimization of Storage Location Assignment and Path Planning in Robotic Mobile Fulfillment Systems. Sustainability; 2021; 13, 5644. [DOI: https://dx.doi.org/10.3390/su13105644]

42. Roshan, K.; Shojaie, A.; Javadi, M. Advanced Allocation Policy in Class-Based Storage to Improve AS/RS Efficiency toward Green Manufacturing. Int. J. Environ. Sci. Technol.; 2019; 16, pp. 5695-5706. [DOI: https://dx.doi.org/10.1007/s13762-018-1921-6]

43. Tappia, E.; Roy, D.; Melacini, M.; De Koster, R. Integrated Storage-Order Picking Systems: Technology, Performance Models, and Design Insights. Eur. J. Oper. Res.; 2019; 274, pp. 947-965. [DOI: https://dx.doi.org/10.1016/j.ejor.2018.10.048]

44. Azadeh, K.; Roy, D.; De Koster, R. Design, Modeling, and Analysis of Vertical Robotic Storage and Retrieval Systems. Transp. Sci.; 2019; 53, pp. 1213-1234. [DOI: https://dx.doi.org/10.1287/trsc.2018.0883]

45. Chen, W.; De Koster, R.; Gong, Y. Analysis and Design of Rack-Climbing Robotic Storage and Retrieval Systems. Transp. Sci.; 2022; [DOI: https://dx.doi.org/10.1287/trsc.2022.1140]

46. Yang, P.; Jin, G.; Duan, G. Modelling and Analysis for Multi-Deep Compact Robotic Mobile Fulfilment System. Int. J. Prod. Res.; 2021; 60, pp. 4727-4742. [DOI: https://dx.doi.org/10.1080/00207543.2021.1936264]

47. D’Antonio, G.; Maddis, M.D.; Bedolla, J.S.; Chiabert, P.; Lombardi, F. Analytical Models for the Evaluation of Deep-Lane Autonomous Vehicle Storage and Retrieval System Performance. Int. J. Adv. Manuf. Technol.; 2018; 94, pp. 1811-1824. [DOI: https://dx.doi.org/10.1007/s00170-017-0313-2]

48. Yener, F.; Yazgan, H.R. Optimal Warehouse Design: Literature Review and Case Study Application. Comput. Ind. Eng.; 2019; 129, pp. 1-13. [DOI: https://dx.doi.org/10.1016/j.cie.2019.01.006]

49. Nicolas, L.; Yannick, F.; Ramzi, H. Order Batching in an Automated Warehouse with Several Vertical Lift Modules: Optimization and Experiments with Real Data. Eur. J. Oper. Res.; 2018; 267, pp. 958-976. [DOI: https://dx.doi.org/10.1016/j.ejor.2017.12.037]

50. Meneghetti, A.; Dal Borgo, E.; Monti, L. Rack Shape and Energy Efficient Operations in Automated Storage and Retrieval Systems. Int. J. Prod. Res.; 2015; 53, pp. 7090-7103. [DOI: https://dx.doi.org/10.1080/00207543.2015.1008107]

51. Hahn-Woernle, P.; Günthner, W.A. Power-Load Management Reduces Energy-Dependent Costs of Multi-Aisle Mini-Load Automated Storage and Retrieval Systems. Int. J. Prod. Res.; 2018; 56, pp. 1269-1285. [DOI: https://dx.doi.org/10.1080/00207543.2017.1395487]

52. Alnahhal, M.; Salah, B.; Ruzayqat, M. An Efficient Approach to Investigate the Tradeoff between Double Handling and Needed Capacity in Automated Distribution Centers. Sustainability; 2022; 14, 7678. [DOI: https://dx.doi.org/10.3390/su14137678]