Fig 2 - uploaded by Goran Dukic
Content may be subject to copyright.
Example of picking route in examined fishbone layout 

Example of picking route in examined fishbone layout 

Source publication
Article
Full-text available
Recently new innovative warehouse layouts are suggested that do not follow usual restrictions of ubiquitous traditional designs. One of them, called fishbone layout, showed potential to reduce travel distances in unit-load warehouses by more then 20%. In manual-pick order-picking systems with case and item picking from multiple locations different...

Contexts in source publication

Context 1
... cross-aisles. The layout with one middle cross aisle is shown in Fig. 1, middle and right. As already said, evaluation of routing policies showed that layouts of order-picking area have significant influence on resulting traveling distances. For a given storage capacity, one can find optimal layout regarding number and length of aisles (Roodbergen and Vis [4], Caron et al. [9]). Results of previous researches showed also that adding one or more cross aisles could benefit the total traveling distances, and that is also possible to find optimal number of cross aisles (Vaughan and Petersen [10]). Although note that adding additional cross aisles increases required storage area (and therefore related costs). The traditional design of warehouse layout is based on a number of unspoken, and unnecessary, assumptions. The two most restrictive are that cross aisles are straight and must meet picking aisles only at right angles, and that picking aisles are straight and are oriented in the same direction. In Gue and Meller [6] authors show that those design assumptions, neither of which is necessary from a construction point of view, limit efficiency and productivity because they require workers to travel longer distances and less-direct routes to retrieve products from racks and deliver them to designated pickup-and-deposit points. In layout that maintains parallel picking aisles, but allows the cross aisle to take different shape, the expected distance to retrieve a single pallet is 8-12% less than in an equivalent traditional design, depending on the dimensions of the warehouse. They named such layout Flying-V layout. Relaxing a second assumption that picking aisles must be parallel, they derived so called fishbone layout. Example of such layout is shown in Fig. 2. The fishbone layout also incorporates the V-shaped cross aisles, with the V extending across the entire warehouse. The picking aisles below the V are horizontal, while the aisles above the V are vertical. The expected travel distance in a fishbone design can be more then 20% less than in a traditional warehouse. Similarly to traditional layouts with cross aisles, these alternative layouts also require a facility 3-5% larger than does the basic traditional layout, which was designed to minimize the footprint of a warehouse. Despite the great potential of new innovative unit-load warehouse designs in reducing traveling distance in pallet picking (single command), the question is what would be the distances of routes for case and item picking from multiple locations in such layouts (multiple command), compared to the traditional layouts. To address this question we tried to analyse routing of order-pickers in fishbone layout. In this paper some preliminary results of analysis are presented. We were restricted on one chosen layout, which is optimal for pallet picking, and the most simplest and used in practice routing method: S-shape. The first problem we encountered was how to define the routing algorithm in fishbone layout. The description for S-shape routing policy in layouts with multiple blocks is as follows (Roodbergen and de Koster [11]). The order-picking route starts at the depot. It goes to the front of the left-most main aisle that contains at least one item. This main aisle is traversed up to and including the block farthest from the depot, that contains at least one item. If the current block contains at least one item, order-picker goes to the left most aisle containing items or to the right most aisle containing items, whichever is the closest. Then goes from one aisle to the next and traverse any aisle containing items entirely. After picking the last item, it returns to the front of the block. If this block contains no items, it traverses the aisle of this block that is closest to the current position. This procedure is repeated for all blocks until the block closest to the depot has been considered. Finally, order-picker returns to the depot. First, it is impossible to say which block in fishbone layout is farthest from the depot, and which is closest to the depot. Second, examining the resulting routes in traditional layout with middle cross aisle (with 2 blocks) and depot located in the middle of the front aisle, we noticed the tendency of above algorithm to create longer routes then expected, due to unnecessary increased across aisle component of routes, as illustrated in Fig 1 middle. Therefore, we suggest modification of algorithm in such situations. The layout is considered as a 3-block warehouse (left down, up, right down). The order-picker starts at the depot, and visits blocks in clockwise manner. In each block aisles containing the items to be picked are also visited in clockwise manner. After picking last item in a block, it goes to the cross aisle in front of the next block, and finally returns to the depot. The resulting route is shown in Fig. 1 right. Same algorithm is easy applicable to the fishbone design, as illustrated in Fig. 2. For analysis we have chosen the layout of 576 locations per layer. Due to the simplicity of distance calculation, dimension of a location is 1x1 meter and the width of all aisles is 2 meters. The traditional layout was with 12 main aisles (total width across aisles 48 meters) and the length of main aisles 24 meters (24 locations per row). With the location of a depot in the middle, it is the optimal layout for single command picking. Comparable fishbone design is shown in Fig. 2. To determine the average traveling distances we used simulation. For a set of orders we generated locations in layout (based on random storage) and calculated traveling distances. The analysis was conducted for 2 order sizes, one relatively small with 10 picks per order, and one large with 30 picks per order. For the purpose of better understanding of routes' nature and comparison, we calculated also both components of travel – within aisle (along main aisles) and across aisle (travel in cross aisles). The results are given in Table 1. As it was expected for examined cases, adding middle cross aisle in traditional layout decreases average routes. For both order sizes density of pick locations in 12 main aisles is not high, and adding middle cross aisle will eliminate some unnecessary travel in main aisles without pick locations. But also note that the percentage of reduction for order size 30 (12.5%) is smaller then reduction for order size 10 (25%). Increasing the order size (increasing the pick density – average distance between picks) there would be the point where adding middle cross aisle is not beneficial. The resulting average routes in fishbone layout are also shorter compared to the basic traditional layout. However, it seems that adding V-shaped cross aisle has smaller potential then adding middle aisle. Across aisle component of average route was even slightly shorter, but reductions of within aisle travels are not as much as for layout with middle cross. That could be explained as follows. First, with fishbone layout we actually have 3 blocks, therefore more aisles then in 2 block layout with middle aisle. Performances of S-shape routing policy are much better in situations with higher number of picks per aisle due to mandatory travel through full length of aisles. Second, fishbone layout creates blocks of aisles with different lengths, with higher probability that pick location is in longer aisles then in shorter aisles. Therefore in total, the order-picker should traverse more aisles where average length of visited aisle is longer then in traditional layout with middle aisle. The required storage area of fishbone layout is also increased in comparison with traditional layout with middle cross aisle. V-shaped cross aisle itself causes some loss of storage area, but main increase is due to two additional rear aisles on left and right side of the layout. In examined situations, the required area for fishbone layout was 16% higher then for basic traditional layout, and 8% higher then for traditional layout with middle aisle. Fishbone layout is with no doubt excellent layout for pallet picking (in our examined case the reduction of single command travel is 13.8%), already implementing in real warehouses. However, in warehouse with case and item picking from multiple locations fishbone layout results in larger routes then traditional layout with straight, right angled cross aisle, at least if S-shape routing policy and random storage are used. But note that longer routes could be due to the nature of S-shape routing policy, which seems not favorable for fishbone layout. More research is needed, regarding other routing policies, storage methods, the shape and size of warehouse, to more completely validate this interesting new layout. Gue and Meller are also working to develop optimal layouts for warehouses that involve case and item picking. Those designs may be totally different. Warehouse designers should be aware of all advantages and disadvantages of different layouts and, depending on the given situation and importance of objectives, choose the most appropriate ...
Context 2
... traveling distances, and that is also possible to find optimal number of cross aisles (Vaughan and Petersen [10]). Although note that adding additional cross aisles increases required storage area (and therefore related costs). The traditional design of warehouse layout is based on a number of unspoken, and unnecessary, assumptions. The two most restrictive are that cross aisles are straight and must meet picking aisles only at right angles, and that picking aisles are straight and are oriented in the same direction. In Gue and Meller [6] authors show that those design assumptions, neither of which is necessary from a construction point of view, limit efficiency and productivity because they require workers to travel longer distances and less-direct routes to retrieve products from racks and deliver them to designated pickup-and-deposit points. In layout that maintains parallel picking aisles, but allows the cross aisle to take different shape, the expected distance to retrieve a single pallet is 8-12% less than in an equivalent traditional design, depending on the dimensions of the warehouse. They named such layout Flying-V layout. Relaxing a second assumption that picking aisles must be parallel, they derived so called fishbone layout. Example of such layout is shown in Fig. 2. The fishbone layout also incorporates the V-shaped cross aisles, with the V extending across the entire warehouse. The picking aisles below the V are horizontal, while the aisles above the V are vertical. The expected travel distance in a fishbone design can be more then 20% less than in a traditional warehouse. Similarly to traditional layouts with cross aisles, these alternative layouts also require a facility 3-5% larger than does the basic traditional layout, which was designed to minimize the footprint of a warehouse. Despite the great potential of new innovative unit-load warehouse designs in reducing traveling distance in pallet picking (single command), the question is what would be the distances of routes for case and item picking from multiple locations in such layouts (multiple command), compared to the traditional layouts. To address this question we tried to analyse routing of order-pickers in fishbone layout. In this paper some preliminary results of analysis are presented. We were restricted on one chosen layout, which is optimal for pallet picking, and the most simplest and used in practice routing method: S-shape. The first problem we encountered was how to define the routing algorithm in fishbone layout. The description for S-shape routing policy in layouts with multiple blocks is as follows (Roodbergen and de Koster [11]). The order-picking route starts at the depot. It goes to the front of the left-most main aisle that contains at least one item. This main aisle is traversed up to and including the block farthest from the depot, that contains at least one item. If the current block contains at least one item, order-picker goes to the left most aisle containing items or to the right most aisle containing items, whichever is the closest. Then goes from one aisle to the next and traverse any aisle containing items entirely. After picking the last item, it returns to the front of the block. If this block contains no items, it traverses the aisle of this block that is closest to the current position. This procedure is repeated for all blocks until the block closest to the depot has been considered. Finally, order-picker returns to the depot. First, it is impossible to say which block in fishbone layout is farthest from the depot, and which is closest to the depot. Second, examining the resulting routes in traditional layout with middle cross aisle (with 2 blocks) and depot located in the middle of the front aisle, we noticed the tendency of above algorithm to create longer routes then expected, due to unnecessary increased across aisle component of routes, as illustrated in Fig 1 middle. Therefore, we suggest modification of algorithm in such situations. The layout is considered as a 3-block warehouse (left down, up, right down). The order-picker starts at the depot, and visits blocks in clockwise manner. In each block aisles containing the items to be picked are also visited in clockwise manner. After picking last item in a block, it goes to the cross aisle in front of the next block, and finally returns to the depot. The resulting route is shown in Fig. 1 right. Same algorithm is easy applicable to the fishbone design, as illustrated in Fig. 2. For analysis we have chosen the layout of 576 locations per layer. Due to the simplicity of distance calculation, dimension of a location is 1x1 meter and the width of all aisles is 2 meters. The traditional layout was with 12 main aisles (total width across aisles 48 meters) and the length of main aisles 24 meters (24 locations per row). With the location of a depot in the middle, it is the optimal layout for single command picking. Comparable fishbone design is shown in Fig. 2. To determine the average traveling distances we used simulation. For a set of orders we generated locations in layout (based on random storage) and calculated traveling distances. The analysis was conducted for 2 order sizes, one relatively small with 10 picks per order, and one large with 30 picks per order. For the purpose of better understanding of routes' nature and comparison, we calculated also both components of travel – within aisle (along main aisles) and across aisle (travel in cross aisles). The results are given in Table 1. As it was expected for examined cases, adding middle cross aisle in traditional layout decreases average routes. For both order sizes density of pick locations in 12 main aisles is not high, and adding middle cross aisle will eliminate some unnecessary travel in main aisles without pick locations. But also note that the percentage of reduction for order size 30 (12.5%) is smaller then reduction for order size 10 (25%). Increasing the order size (increasing the pick density – average distance between picks) there would be the point where adding middle cross aisle is not beneficial. The resulting average routes in fishbone layout are also shorter compared to the basic traditional layout. However, it seems that adding V-shaped cross aisle has smaller potential then adding middle aisle. Across aisle component of average route was even slightly shorter, but reductions of within aisle travels are not as much as for layout with middle cross. That could be explained as follows. First, with fishbone layout we actually have 3 blocks, therefore more aisles then in 2 block layout with middle aisle. Performances of S-shape routing policy are much better in situations with higher number of picks per aisle due to mandatory travel through full length of aisles. Second, fishbone layout creates blocks of aisles with different lengths, with higher probability that pick location is in longer aisles then in shorter aisles. Therefore in total, the order-picker should traverse more aisles where average length of visited aisle is longer then in traditional layout with middle aisle. The required storage area of fishbone layout is also increased in comparison with traditional layout with middle cross aisle. V-shaped cross aisle itself causes some loss of storage area, but main increase is due to two additional rear aisles on left and right side of the layout. In examined situations, the required area for fishbone layout was 16% higher then for basic traditional layout, and 8% higher then for traditional layout with middle aisle. Fishbone layout is with no doubt excellent layout for pallet picking (in our examined case the reduction of single command travel is 13.8%), already implementing in real warehouses. However, in warehouse with case and item picking from multiple locations fishbone layout results in larger routes then traditional layout with straight, right angled cross aisle, at least if S-shape routing policy and random storage are used. But note that longer routes could be due to the nature of S-shape routing policy, which seems not favorable for fishbone layout. More research is needed, regarding other routing policies, storage methods, the shape and size of warehouse, to more completely validate this interesting new layout. Gue and Meller are also working to develop optimal layouts for warehouses that involve case and item picking. Those designs may be totally different. Warehouse designers should be aware of all advantages and disadvantages of different layouts and, depending on the given situation and importance of objectives, choose the most appropriate ...
Context 3
... negligible influence of layout on performances of particular method or mix of methods. All papers regarding analysis of order-picking methods in manual-pick order-picking systems imply traditional layouts. Just recently engineering professors Russell Meller and Kevin Gue proposed radically new, innovative warehouse layouts that could reduce retrieval times in pallet picking (Gue and Meller [6]). In this paper we present the preliminary results of analysis of the simplest but also most common in practice routing policy for picking from multiple locations, in so called "fishbone" innovative layout, in comparison with performances in traditional layouts. The paper is divided as follows. In Section 2 we give brief description of developed routing methods. In Section 3 we present the traditional and innovative layouts, with review of former research results regarding influence of layout on traveling distances in order-picking. In Section 4 examined situation and results of analysis are presented, while conclusions are drawn in Section 5. Routing of order-picker concerns the movement of the order picker from location to location to retrieve products. The objective of routing policies is to sequence the items on the pick list to ensure a good route through the picking area – as short as possible. The problem of routing order pickers in a warehouse is actually a special case of the Travelling Salesman Problem. The order picker starts at the depot, has to visit all pick locations and finally has to return to the depot. For the type of warehouse shown in Fig. 1 left, Ratliff and Rosenthal [7] developed an algorithm that results in a shortest possible, thus optimal route, while Roodbergen and de Koster [8] developed an algorithm for shortest route in warehouses with 2 blocks (with additional cross aisle in the middle), type shown in Fig. 1 middle and right. Besides the objective of short routes, there are other considerations. Order-picker has to execute the route, so it should be easy to understand and follow, which also could lead to enhanced productivity. This is probably the reason while most warehouses use heuristic routing policies. There are several heuristic routing methods (policies) developed and used in practice. The simplest routing heuristic is the S-shape policy. When this method is used, the order picker enters every aisle where an item has to be picked and traverses the entire aisle. Aisles where nothing has to be picked are skipped. An exception is made for the last aisle visited, in case the number of aisles to be visited is odd. In that case a return travel is performed in the last aisle visited. For descriptions of other routing policies and their evaluations please refer to the literature listed in de Koster et al. [3]. Traditional warehouse/order-picking area layouts are layouts we could find today in majority of warehouses. The basic form is with parallel aisles, a central depot (pick up/delivery point), and two possibilities for changing aisles, at the front and at the rear of warehouse, shown in Fig. 1 left. Modifications of this basic form are usually with adding one or more additional cross aisles. In this case we refer to a layout with multiple cross-aisles. The layout with one middle cross aisle is shown in Fig. 1, middle and right. As already said, evaluation of routing policies showed that layouts of order-picking area have significant influence on resulting traveling distances. For a given storage capacity, one can find optimal layout regarding number and length of aisles (Roodbergen and Vis [4], Caron et al. [9]). Results of previous researches showed also that adding one or more cross aisles could benefit the total traveling distances, and that is also possible to find optimal number of cross aisles (Vaughan and Petersen [10]). Although note that adding additional cross aisles increases required storage area (and therefore related costs). The traditional design of warehouse layout is based on a number of unspoken, and unnecessary, assumptions. The two most restrictive are that cross aisles are straight and must meet picking aisles only at right angles, and that picking aisles are straight and are oriented in the same direction. In Gue and Meller [6] authors show that those design assumptions, neither of which is necessary from a construction point of view, limit efficiency and productivity because they require workers to travel longer distances and less-direct routes to retrieve products from racks and deliver them to designated pickup-and-deposit points. In layout that maintains parallel picking aisles, but allows the cross aisle to take different shape, the expected distance to retrieve a single pallet is 8-12% less than in an equivalent traditional design, depending on the dimensions of the warehouse. They named such layout Flying-V layout. Relaxing a second assumption that picking aisles must be parallel, they derived so called fishbone layout. Example of such layout is shown in Fig. 2. The fishbone layout also incorporates the V-shaped cross aisles, with the V extending across the entire warehouse. The picking aisles below the V are horizontal, while the aisles above the V are vertical. The expected travel distance in a fishbone design can be more then 20% less than in a traditional warehouse. Similarly to traditional layouts with cross aisles, these alternative layouts also require a facility 3-5% larger than does the basic traditional layout, which was designed to minimize the footprint of a warehouse. Despite the great potential of new innovative unit-load warehouse designs in reducing traveling distance in pallet picking (single command), the question is what would be the distances of routes for case and item picking from multiple locations in such layouts (multiple command), compared to the traditional layouts. To address this question we tried to analyse routing of order-pickers in fishbone layout. In this paper some preliminary results of analysis are presented. We were restricted on one chosen layout, which is optimal for pallet picking, and the most simplest and used in practice routing method: S-shape. The first problem we encountered was how to define the routing algorithm in fishbone layout. The description for S-shape routing policy in layouts with multiple blocks is as follows (Roodbergen and de Koster [11]). The order-picking route starts at the depot. It goes to the front of the left-most main aisle that contains at least one item. This main aisle is traversed up to and including the block farthest from the depot, that contains at least one item. If the current block contains at least one item, order-picker goes to the left most aisle containing items or to the right most aisle containing items, whichever is the closest. Then goes from one aisle to the next and traverse any aisle containing items entirely. After picking the last item, it returns to the front of the block. If this block contains no items, it traverses the aisle of this block that is closest to the current position. This procedure is repeated for all blocks until the block closest to the depot has been considered. Finally, order-picker returns to the depot. First, it is impossible to say which block in fishbone layout is farthest from the depot, and which is closest to the depot. Second, examining the resulting routes in traditional layout with middle cross aisle (with 2 blocks) and depot located in the middle of the front aisle, we noticed the tendency of above algorithm to create longer routes then expected, due to unnecessary increased across aisle component of routes, as illustrated in Fig 1 middle. Therefore, we suggest modification of algorithm in such situations. The layout is considered as a 3-block warehouse (left down, up, right down). The order-picker starts at the depot, and visits blocks in clockwise manner. In each block aisles containing the items to be picked are also visited in clockwise manner. After picking last item in a block, it goes to the cross aisle in front of the next block, and finally returns to the depot. The resulting route is shown in Fig. 1 right. Same algorithm is easy applicable to the fishbone design, as illustrated in Fig. 2. For analysis we have chosen the layout of 576 locations per layer. Due to the simplicity of distance calculation, dimension of a location is 1x1 meter and the width of all aisles is 2 meters. The traditional layout was with 12 main aisles (total width across aisles 48 meters) and the length of main aisles 24 meters (24 locations per row). With the location of a depot in the middle, it is the optimal layout for single command picking. Comparable fishbone design is shown in Fig. 2. To determine the average traveling distances we used simulation. For a set of orders we generated locations in layout (based on random storage) and calculated traveling distances. The analysis was conducted for 2 order sizes, one relatively small with 10 picks per order, and one large with 30 picks per order. For the purpose of better understanding of routes' nature and comparison, we calculated also both components of travel – within aisle (along main aisles) and across aisle (travel in cross aisles). The results are given in Table 1. As it was expected for examined cases, adding middle cross aisle in traditional layout decreases average routes. For both order sizes density of pick locations in 12 main aisles is not high, and adding middle cross aisle will eliminate some unnecessary travel in main aisles without pick locations. But also note that the percentage of reduction for order size 30 (12.5%) is smaller then reduction for order size 10 (25%). Increasing the order size (increasing the pick density – average distance between picks) there would be the point where adding middle cross aisle is not beneficial. The resulting average routes in fishbone layout are also shorter compared to the basic traditional layout. However, it seems that adding V-shaped cross ...

Similar publications

Article
Full-text available
Information-Centric Networks (ICN) are network architec-tures aiming at decoupling locator and identifier of contents, the primary objective of communication. Being the most popular proposal of the family of ICN, NDN specifically adopts pervasive network caches and name-based forwarding at network layer. It has been widely assumed these attributes...
Article
Full-text available
This paper presents an approach to estimate bus route and network travel times using micro-simulation. This can be used in predicting the effectiveness of bus route designs using some network traffic measures or indicators. The used indicators are average network traffic intensity, posted speeds, route length, frequency of bus operation, and averag...

Citations

... Regarding the use of optimization algorithms to sort warehouse goods, Steffey [10], Huang et al. [11], Opetuk and Dukic [12], Zhou et al. [13], Pohl et al. [14], Xu and Hu [15], and Ardjmand et al. [16] used fishbone layout and artificial fish school algorithm and improved cat group algorithm, respectively, and used genetic algorithm analysis and comparison to find the picking path and picking of the warehouse. Regarding the dispatching optimization plan of the cargo trolley, Venkitasubramony and Adil [17] and Le and Degui [18] (2013) constructed a warehouse location allocation model based on layout Fishbone, used a hybrid algorithm combining genetic algorithm and ant colony algorithm, and improved layout Fishbone methods to optimize the allocation of many locations. ...
Article
Full-text available
To study the genetic algorithm, this paper solves the problem of shop scheduling under the premise of layout Flying−V. Firstly, double-layer coding is used for optimization. When calculating fitness, the time to return to the mouth P&D approaches the optimum through the greedy idea. Individual screening is carried out through the roulette method. Different crossover and genetic operators are used for different coding layers. Through thinking of elitism and catastrophe and the immigration operator to ensure the diversity of the algorithm in the calculation process, it can achieve the recommendation of the number of cars to control the cost. The stability of the algorithm is good. It can recommend a better picking sequence and number of carts for various types of picking problems.
... Picking Order is a process of taking goods starting from take the picking slip as written instructions for taking goods until handover the checker. Order Picking is the process of picking up goods from a shelf location based on customer demand [7]. Picking Order is a function which is critical for managing and operating distribution warehouses efficiently [8] Order Picking is the most expensive activity in warehousing and can reach 55% of the total cost of warehousing operations, so it is considered a top priority in increasing productivity, even reaching 65% of total warehouse operating costs [9]. ...
... For manual pick-picking systems, travel time increases related with mileage, so the mileage is often the main goal in the planning and optimizing the warehouse. According to Roodbergen [10] the method for efficient picking orders that combine travel times can be categorized into 3 groups: routing, storage and batching Routing method for determining the direction and route of travel and used for mileage, storage method is used to allocate goods to the location based on certain rules, and order batching is grouping customer orders into one picking order [7]. A good routing strategy can reduce travel time as much as 30% [15]. ...
Article
Full-text available
Spare parts are one of the important pillars in the after-sales service of automotive business. Customers will satisfied and comfortable if the availability of spare parts is guaranteed. Spare Part Center is one of function to support unit sales and as well as profit-oriented, so the accuracy and speed of spare part acceptance by the customer is an important key to winning the competition. Order Picking is one of the supply chain processes that play a role in warehouse operations to meet customer needs. Order Picking is the most expensive activity in warehousing and can reach 55% of the total cost of warehousing operations, so it is considered a top priority in increasing productivity, even reaching 65% of total warehouse operating costs. The purpose of this research is to increase productivity in the process of picking order through reduction of processing time. Increased productivity is done by improving the working method of the picking process. From the result the comparing, the method by zone requires less total picking time (193.712 seconds) than by routing (249.559 seconds) decreased 55.85 second time, in other words, an increase of 22.38%. With the Visual Stream Mapping (VSM) in this research can reduce to travel time, it means that the total distance traveled is small than the current method. The impact from VSM approach will eliminate time for preparation of 1.960 seconds, and take empty trolley of 200 seconds. In this case some of traveling non-value adding for customer, so this is should be taken off from process picking because it is a waste.
... In (Dukic & Opetuk, 2008) the fishbone layout was tested for manual-pick order-picking systems with different routing policies. They show that with the most common routing polices, the fishbone layout increases the distances of the travel. ...
Article
Full-text available
The purpose of this paper is to study the Fishbone Warehouse layout. Fishbone layouts are non-traditional designs for warehouses that have shown to be more efficient under unit-load operations. This paper presents an analytical study of the design from an optimisation point of view. The authors gain an insight into the characteristics of the layout and present alternatives for different situations of interest in industrial environments. Finally, the authors compare the performance of Fishbone and traditional designs; presenting a formal conception of the expected savings. The authors focus on finding optimal conditions for the most important characteristic of the Fishbone design, which is the slope of the diagonal cross-aisle. The design decisions are modelled as a non-linear optimisation problem and solved using both numerical and exact methods, depending on the most convenient analysis. In addition, the robustness of the results is presented and their practical implications discussed.
... Despite the great potential of new innovative unit-load warehouse designs in reducing traveling distance in pallet picking (single command), the question is what would be the distances of routes for case and item picking from multiple locations in such layouts (multiple command), compared to the traditional layouts. To address this question, an analysis was done in [25] with the simplest and commonly used in practice S-shape routing method, and also extended for this paper with more complex Composite routing method. Figure 5 illustrates one example of a routing using S-shape method modified to be adapted for analyzed fishbone layout. ...
Article
Full-text available
Green supply chain management is a concept that is gaining popularity all over the world. Besides, it is a way to demonstrate commitment to sustainability and to be fully adopted by the organizations it should contribute to better economic performances and competitiveness. Recently there have been many incentives for more sustainable warehousing in supply chains. In order to improve efficiency of order-picking in warehouses, there are many methods, models and technologies developed and used. This paper presents, after a brief overview of green supply chain management, an overview of order-picking methods and technologies and their potentials in improving order-picking efficiency, based mainly on reducing traveling distances. In this way energy consumption is reduced, influencing also greening of warehousing too.
... These types of results started discussions on how applicable the design was and what would happen if some of these assumptions were relaxed. Dukic and Opetuk's (2008) interest in this topic came from the results presented by Meller and Gue (2006). As mentioned above, Meller and Gue (2006) used the assumption of only unit load picks where Dukic and Opetuk (2008) wanted to relax this assumption to see if the Fishbone Racking Model was optimal for an order picking distribution center. ...
... Dukic and Opetuk's (2008) interest in this topic came from the results presented by Meller and Gue (2006). As mentioned above, Meller and Gue (2006) used the assumption of only unit load picks where Dukic and Opetuk (2008) wanted to relax this assumption to see if the Fishbone Racking Model was optimal for an order picking distribution center. Order picking is the process of going to multiple locations within the storage area to complete an order. ...
... There are multiple policies for determining the optimal pick path; however, the issue with these policies is that they were developed for the traditional racking layout. Dukic and Opetuk (2008) developed three different layouts to compare. The three different layouts were: a traditional layout without a cross aisle, a traditional layout with a cross aisle and the third was the Fishbone Racking Model. ...
... • The papers of Dukie ,Opetuk [60] and Pohi et al [61] deal with mathematical modelling for fishbone aisle warehouse layout. ...
Article
Full-text available
Warehouse design and operation plays an integral role in the whole supply chain system. For this reason, many researchers have dealt with these types of problems not only by mathematically modelled but also by introduced other kinds of solutions such as artificial intelligence. On our paper, we present these two types of solutions (mathematical models and technological solutions) and the types of warehouse layout problems. Our goal is to offer a more structured view for the important problem of warehouse layout problem.
Article
A recent trend in the layout design of unit-load warehouses is the application of layouts without conventional parallel pick aisles and straight middle aisles. Two examples for such designs are flying-V and fishbone designs for single- and dual-command operations. In this study, it is shown that the multi-item order picking problem can be solved in polynomial time for both fishbone and flying-V layouts. These two designs are compared with the traditional parallel-aisle design under the case of multi-item pick lists. Simple heuristics are proposed for fishbone layouts that are inspired by those put forward for parallel-aisle warehouses and it is experimentally shown that a modification of the aisle-by-aisle heuristic produces good results compared with other modified S-shape and largest gap heuristics when items have uniform demand. Computational experiments are performed in order to compare the performances of fishbone and traditional layouts under optimal routing and it is shown that a fishbone design can obtain improvements of around 20% over parallel-aisle design in single-command operations but can perform as high as around 30% worse than an equivalent parallel-aisle layout as the size of the pick list increases. The sensitivity of the results to varying demand skewness levels when volume-based storage is applied is tested and it is shown that unlike the single- and dual-command cases, a fishbone design performs better compared to a traditional design under highly skewed demand as opposed to uniform demand.