Improving the Multi-Brand Channel Distribution of a Fashion Retailer

Available online at www.sciencedirect.com

ScienceDirect ScienceDirect Procedia Manufacturing 00 (2018) 000–000

Available Availableonline onlineatatwww.sciencedirect.com www.sciencedirect.com Procedia Manufacturing 00 (2018) 000–000

ScienceDirect ScienceDirect 

www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia

Procedia Manufacturing 17 (2018) 655–662 Procedia Manufacturing 00 (2017) 000–000 www.elsevier.com/locate/procedia

28th International Conference on Flexible Automation and Intelligent Manufacturing (FAIM2018), 11-14, 2018, Columbus, USA Manufacturing 28th International ConferenceJune on Flexible Automation and OH, Intelligent (FAIM2018), June 11-14, 2018, Columbus, OH, USA

Improving the Multi-Brand Channel Distribution of a Fashion Retailer Improving Multi-Brand Channel Distribution of a Fashion Retailer 1 1,2 1 MESIC 2017,1 28-30 June Manufacturingthe Engineering Society International Conference 2017, Tiago Silva , Teresa Pereira , Luís Pinto Ferreira , F. J. G. Silva 2017, Vigo1,2(Pontevedra), Spain 1 1 1 , Teresa Pereira Luís Pinto Ferreira F. 4200-072 J. G. Silva ISEP –Tiago School ofSilva Engineering, Polytechnic of Porto, R.,Dr. Antº Bernardino de Almeida,,431, Porto, Portugal

1

Costing models for capacity optimization in Industry 4.0: Trade-off between used capacity and operational efficiency 2

Research Center of Mechanical Engineering (CIDEM), School of Engineering, Polytechnic of Porto, 4200-072 Porto, Portugal 1 ISEP – School of Engineering, Polytechnic of Porto, R. Dr. Antº Bernardino de Almeida, 431, 4200-072 Porto, Portugal 2 Research Center of Mechanical Engineering (CIDEM), School of Engineering, Polytechnic of Porto, 4200-072 Porto, Portugal

Abstract

A. Santanaa, P. Afonsoa,*, A. Zaninb, R. Wernkeb

Abstract a University of Minho, Portugal One has seen exponential growth in the number of clients4800-058 and in Guimarães, the quantities ordered in the fashion retailing multi-brand b 89809-000 SC, Brazil channel. It has, therefore, become essential toUnochapecó, improve the channel'sChapecó, distribution process in order to meet the customers’ orders in One has seen exponential growth in themanner, number thus of clients and in thethe quantities in theupon fashion multi-brand the shortest time, and in a cost-effective complying with deliveryordered terms agreed withretailing the market. To this channel. It has, therefore, become essential to improve the channel's distribution process in order to meet the customers’ in end, one studied aspects such as the mapping of the supply flow process, the occupation of space and the spaghetti-dashorders diagram the shortest time, and in a cost-effective manner, thus these, complying withanalyzed the delivery terms agreed with thecycle market. To takt this of four current distribution process activities. Besides one also the calculation of upon productivity, times, end, studied aspects such as the mapping of the the purpose supply flow process, athe occupation of space and theperformance. spaghetti-dash Abstract time,one as well as the service level designed, with of preparing system to evaluate company In diagram addition of these four current process these, one analyses also analyzed the to calculation of productivity, times, takt to studies,distribution one resorted to the activities. ABC and Besides SWOT customers’ in order develop the improvement cycle proposal, which time,characterized as the wellconcept as theby: service level designed, with the of preparing system to evaluate performance. In addition was (a) in the layout, (b) purpose improvements in will the asupply flow, (c)toimplementation of gravity carriers, as Under of changes "Industry 4.0", production processes be pushed becompany increasingly interconnected, to these studies, one resorted toofthe ABC and SWOT analyses in order todeveloped develop the proposal, for which well as more ergonomic forms transport, and (d) thecustomers’ use of much computer applications in improvement Visualcapacity Basic language the information based on a real time basis and, necessarily, more efficient. In this context, optimization was characterized by:Based (a) changes in the layout, (b) improvements in thethe supply flow, (c) implementation of gravity carriers, as distribution process. on aim this proposal, onemaximization, succeeded in increasing amount sorted out by the distributor in anand eight-hour goes beyond the traditional of capacity contributing also for organization’s profitability value. well as more ergonomic forms of transport, and (d) the use of computer applications developed in Visual Basic language for the shift to 294 articles (11,23%). Cycle time was reduced from 0,015 minutes/article to 0,013 minutes/article (13,33%), which Indeed, lean management andproposal, continuous improvement approaches suggest capacity instead of distribution process. Based of on articles this in increasing amount sorted out byspace theoptimization distributor in anto eight-hour allows for the segregation in timeone forsucceeded the next collections. Inthe addition, the occupied was reduced 47 m2 on shift to 294 articles (11,23%). Cycle time was reduced from 0,015 minutes/article to 0,013 minutes/article (13,33%), which maximization. The study of capacity optimization and costing models is an important research topic that deserves average per collection (1,39%), which is translated into a reduction of 1 498 468 meters (23,34%) in the average distance covered on allows for the segregation ofthe articles in time for the was next reduced, collections. In This addition, thepresents occupied space was reduced to 47 m2The contributions from both the practical and theoretical perspectives. paper and discusses per collection. Furthermore, number of workers on average, by five employees (12,82%) pera mathematical collection. average per collection (1,39%), which is translated into a reduction of 1 498 468 meters (23,34%) in the average distance covered model capacity based different (ABC and A generic has been storage for capacity of themanagement finished product wason also increasedcosting by 535models boxes (11,30%). TheTDABC). total investment neededmodel to achieve these per collection. Furthermore, the 23 number of workers reduced, average,will by only five employees (12,82%) per in collection. The changes is established as being 754,42 yet, thewas payback timeon involved bethe six maximization months, resulting a cumulative developed and it was used to analyze idle€;capacity and to design strategies towards of organization’s storageofcapacity of €the finished product was also increased by 535 boxes (11,30%). The total investment needed to achieve these profit 84 504,23 by the end of the fall/winter 2020 collection. value. trade-off as capacity is highlighted and it isresulting shown inthat capacity changesThe is established being 23maximization 754,42 €; yet, vs theoperational payback timeefficiency involved will only be six months, a cumulative optimization might hide operational inefficiency. profit of 84 504,23 € by the end of the fall/winter 2020 collection. © © 2017 2018 The The Authors. Authors. Published Published by by Elsevier B.V. B.V. © 2018 The Authors. Published by B.V. committee Peer-review under responsibility of Elsevier the of the Manufacturing Engineering Society International Conference This is an open access article under the scientific CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/) © 2018 The Authors. by Elsevier B.V. This is an open accessPublished article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/) 2017. Peer-review under responsibility of the scientific committee of the 28th Flexible Automation and Intelligent Manufacturing Peer-review under responsibility committee the 28th Flexible Automation and Intelligent Manufacturing This is an open access article underofthethe CCscientific BY-NC-ND license of (https://creativecommons.org/licenses/by-nc-nd/4.0/) (FAIM2018) Conference. (FAIM2018) Conference. Peer-review under responsibility of the scientific committee of the 28thOperational Flexible Automation Keywords: Cost Models; ABC; TDABC; Capacity Management; Idle Capacity; Efficiency and Intelligent Manufacturing (FAIM2018) Conference. Keywords: Improvement; Layout; Logistics; Picking; Productivity; Supply flows. Keywords: Improvement; Layout; Logistics; Picking; Productivity; Supply flows.

1. Introduction

2351-9789 © 2018 The Authors. Published by Elsevier B.V. idle article capacity a fundamental for companies and their management of extreme importance ThisThe is ancost openof access underisthe CC BY-NC-NDinformation license (https://creativecommons.org/licenses/by-nc-nd/4.0/) 2351-9789 ©under 2018responsibility The Authors. of Published by Elsevier Peer-review theIn scientific committee of the 28th Flexiblecapacity Automation Intelligent potential Manufacturing in modern production systems. general, it isB.V. defined as unused orand production and(FAIM2018) can be measured This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/) Conference. in several ways: tons of production, available hours of manufacturing, etc. The management of the idle capacity Peer-review under responsibility of the scientific committee of the 28th Flexible Automation and Intelligent Manufacturing (FAIM2018) * Paulo Afonso. Tel.: +351 253 510 761; fax: +351 253 604 741 Conference. E-mail address: [email protected] 2351-9789 Published by Elsevier B.V. B.V. 2351-9789©©2017 2018The TheAuthors. Authors. Published by Elsevier Peer-review underaccess responsibility the scientific committee oflicense the Manufacturing Engineering Society International Conference 2017. This is an open article of under the CC BY-NC-ND (http://creativecommons.org/licenses/by-nc-nd/3.0/) Peer-review under responsibility of the scientific committee of the 28th Flexible Automation and Intelligent Manufacturing (FAIM2018) Conference. 10.1016/j.promfg.2018.10.114

656 2

Tiago Silva et al. / Procedia Manufacturing 17 (2018) 655–662 Author name / Procedia Manufacturing 00 (2018) 000–000

1. Introduction The efficiency and effectiveness of logistic operations are of crucial importance when determining the level of an organization’s competitiveness today [1]. The globalization of the market, great competition, short product lifespans, high productivity and the reduction in time-to-market are all factors which condition operation performance [2]. Gaining a competitive edge is directly related to the appropriate choice of warehouse layout, since this produces immediate effects both in the initial capital invested as well as in operational costs [3]. On the shop floor, between 55% and 75% of the total warehouse costs are generated by the most time-consuming activity in the area of logistics, which is known as order-picking [4]. The selection of suitable equipment for storage, transportation, order picking and sorting essentially determines the level of warehouse automation, and this ultimately influences the performance achieved [5]. The implementation of a WMS (Warehouse Management System) to support warehouse management allows for an increase in efficiency as well as a reduction in costs, when compared to a manual management system [6]. The work described in this article was developed at a fashion retail company. Its purpose was to improve the distribution process of a multi-brand channel so that, in the period between the FW17 (Fall/Winter 2017) and FW20 (Fall/Winter 2020) collections, one could meet the customers’ orders in the shortest time and at the lowest cost possible, thus complying with the due date for delivery agreed upon with the end consumer. The article is divided into six sections: section 1 consists of the introduction; section 2 presents a review of literature dealing with the analysis and optimization of warehouses; section 3 describes the methodology used to carry out this research work; section 4 presents all the practical work developed at the company; section 5 contains the results obtained: and section 6 consists of the conclusion of the study. 2. Literature Review There are countless studies in literature pertaining to the subject of warehouse analysis and optimization. These deal with the identification and elimination of waste, with the subsequent adoption of the most suitable strategies of distribution for each organization, thus enabling a significant increase in logistic performance. In the case study of a warehouse belonging to a tea producer, an algorithm was developed to generate work orders according to the combinations of the picking and packaging operations. This led to the elimination of extra storage and to a reduction in operation time [7]. In a warehouse owned by a producer of interior design objects, it was demonstrated that logistic performance could be improved by means of the Warehouse Management System, in which VSM (Value Stream Mapping) allowed for the identification of the activity area where waste was being generated. In addition, Genba Shikumi allowed for the categorization of the weight of identified waste [8]. In another study, which considered a rectangular warehouse with a rack system, an algorithm was developed to support decision-making for a warehouse layout. This enabled minimizing space and reducing the costs associated to material flow [9]. In line with the results achieved in a study undertaken of German supermarket distribution centers, it was concluded that, if the picking process time is less than that of the mobile rack displacement process, it is advantageous to use the computational results suggested in order to minimize the space taken up by the racks [10]. The use of bar codes for the two-bin system, used in a logistics warehouse at an electronic components company, resulted in the elimination of unnecessary movements, as well as in a reduction of waiting time and the distance covered [11]. In the case of a company in the tire industry, the implementation of the RFID (Radio Frequency Identification) system led to a reduction in the costs of logistics and to better service rendered to the end customer [12]. An ABC analysis of product rotation at a company in the automotive sector encouraged the change in material management from FIFO (First-in First-out) to FEFO (First-expired First-out). A new layout was also determined, and the calculation for the number of bins required enhanced logistic performance [13]. The most time-consuming activity in logistics, which is termed “picking”, consists of removing articles ordered by the customer from the warehouse stock. This must be carried out at the right moment and in the required quantity, with the subsequent delivery of the final product to the customer in the best possible conditions [14]. The mapping of processes by means of the spaghetti diagram, as well as through the business process model and notation, contributes to the efficient and effective identification and elimination of non-value adding activities. The business process model and notation consists of a notation method based on flowcharts used to define business processes. Namely, it combines graphic features and support data, with the ultimate aim of providing further details of the process at hand [15]. The spaghetti diagram involves plotting lines on the gemba layout, which reproduce the observation of the flow of material movements, as well as that of information and human resources [16]. The principal performance metrics or the KPI (Key Performance Indicator)



Tiago Silva et al. / Procedia Manufacturing 17 (2018) 655–662 Author name / Procedia Manufacturing 00 (2018) 000–000

657 3

used in the development of this study, and which will support the evaluation of the process are: P (Productivity), CT (Cycle Time) and TK (Takt Time). Productivity measures the product’s capacity to transform input (resources) into output (products); namely, productivity consists of maximizing the use of resources, offering the customer a product at the lowest possible cost [17]. Cycle time measures the time between successive articles; it is defined by the lengthiest activity, known as the bottleneck, which conditions production rate [18]. Takt time is the term used to refer to the time cycle which meets customer requirements and is the heartbeat of distribution [19]. 3. Methodology The methodology used to undertake this study was divided into various stages. The first of these consisted of a review of literature, which was supported by scientific articles dealing with the field of warehouse analysis and optimization. The purpose here was to ensure a coherent support for the empirical research presented. During the second stage, one proceeded with the activity of mapping the process, as well as that of the supply flow and the occupied space. Spaghetti diagrams were also drawn up, and a calculation of the performance measures was carried out to allow for an analysis of the distribution process used in the multi-brand channel. The critical points and potentialities were thus identified. It was in the third stage that one drafted the proposals for improvement in the distribution process of the multi-brand channel. In the fourth and last stage, one proceeded with the implementation of the suggested proposals, which was supported by the Pan, Do, Check, Act (PDCA) methodology. 4. Improvements in the multi-brand channel distribution process The distribution process used by the multi-brand channel of a fashion retail company is organized in four activities, which are designated as: multi-brand pick by line; storage of the finished product; storage of raw materials; and product expedition. 4.1. Mapping of the current distribution process The logistics warehouse at the company (see fig. 1) occupies an area of 8 700 m2, in which 2 497 m2 are associated to the channel studied. While 1 125 m2 are allocated to the multi-brand pick by line activity, 750 m2 are used for the storage of the finished product, 422 m2 are taken up by raw material storage and 200 m2 are used for product expedition. (1) Raw material warehouse

(2) Multi-brand pick by line

(3) Storage of finished product

(4) Product expedition

Up

4

3 2 Up

Up

X2

1 25 m2

PASSAGEWAY FOR MACHINES

PASSAGEWAY FOR PEOPLE

STACKER

COMPUTER

AMERICAN PALLET

RACK

STRUCTURAL PILLAR

Fig. 1. Layout of the logistics warehouse at a fashion retail company (current distribution process).

UNITS: METERS 8 700 m2

Tiago Silva et al. / Procedia Manufacturing 17 (2018) 655–662 Author name / Procedia Manufacturing 00 (2018) 000–000

658 4

4.2 Evaluation of the performance of the current distribution process An ABC analysis of the packaging process allowed one to conclude that, for the spring/summer 2017 collection, 30 661 boxes were packaged for 2 069 customers. Of these, 1 to 15 boxes were packaged for 1 526 customers, between 16 and 37 boxes were packaged for 399 customers, while 38 to 1 031 boxes were packaged for 144 customers. Thus, 6,96% of the customers are considered to be class A, 19,28% are considered to be class B and 73,76% are considered as class C. One also analyzed the product family for which the largest number of article displacements occurs in the raw materials warehouse in order to execute the multi-brand pick by line activity. One was thus able to determine the number of pallets moved (see table 1): the trousers range is responsible for most of the pallet movements, and is followed by the family of seamless items, jackets and coats and, lastly, shirts and knitwear. Table 1. Number of pallets moved for each product family. Product family

Total nº of boxes

Nº of boxes/pallet

Total nº of pallets

Trousers

11 148

30

372

Seamless items

9 120

25

365

Knitwear

1 542

25

62

Jackets and coats

5 695

20

285

Shirts

3 717

30

124

Total

31 222

26

1 208

272 271 273 272

271 270

270 269

269 268

268 267

267 266

266 265

265 264

264 263

263 262

251 250 252 251 253 252 254 253 255 254 256 255 257 256 258 257 259 258 260 259

210 209 211 210 212 211 213 212 214 213 215 214 216 215 217 216

201 200 202 201 203 202 204 203 205 204 206 205 207 206

272 273

216 217

207 208

208 207

271 272

260 261

270 271

259 260

269 270

258 259

268 269

257 258

267 268

256 257

266 267

255 256

265 266

254 255

215 216

206 207

214 215

205 206

213 214

204 205

212 213

203 204

211 212

202 203

49.961 meters/pallet

210 211

201 202

(1) Raw material warehouse

209 210

200 201

Fechar Caixa

264 265

253 254

1.891 meters/article

263 264

252 253

(2) Multi-brand pick by line Fechar Caixa

262 263

251 252

24.407 meters/box

250 251

(3) Storage of finished product 25.961 meters/box

(2) Multi-brand pick by line

261 260

The results obtained from the spaghetti diagram, which plots the activities associated to the current distribution process of the multi-brand channel, allowed one to determine the distance covered per article, per box or per pallet in each of the activities (see fig. 2).

(4) Product expedition 65.100 meters/pallet

Fig. 2. Representation of the spaghetti diagrams for the activities of the current distribution process.

While 5 164 887 meters were covered for the spring/summer 2017 collection, it is expected that 5 632 942 meters will be covered for the fall/winter 2017 collection. The distance which will be covered for the spring/summer 2018 collection is estimated at 5 702 230 meters while, for the fall/winter 2018 collection, it is expected to be 6 210 981



Tiago Silva et al. / Procedia Manufacturing 17 (2018) 655–662 Author name / Procedia Manufacturing 00 (2018) 000–000

659 5

meters. The number of meters covered for the spring/summer 2019 collection is foreseen to be 6 287 542 meters, and for the fall/winter 2019 collection it is estimated to be 6 848 483 meters. Finally, the expected distance covered for the spring/summer 2020 collection is 6 828 797 meters, while the fall/winter 2020 collection will be 7 438 012 meters. Takt time was established for each collection by combining two types of data. Firstly, one considered the time available for the distribution of the articles ordered by customers from each collection, which was determined as being 31 680 minutes. Subsequently, one also took the expected quantity ordered into account. When comparing these aspects to the cycle time of 0,015 minutes/article, it was concluded that under current operation conditions, it will be impossible to fill the orders placed by customers within the agreed time with the end consumer for the collections of spring/summer 2019, spring/summer 2020 and fall/winter 2020 (see fig. 3).

Time (minutes/article)

Cycle time vs. Takt time 0,020 0,018 0,016 0,014 0,012 0,010 0,008 0,006 0,004 0,002 0,000

Takt time (minutes/article) Cycle time (minutes/article)

FW17

SS18

FW18

SS19

FW19

SS20

FW20

Collection Fig. 3. Comparison of takt time per collection with the cycle time of the current distribution process.

In order to determine the area taken up by the four activities in the current distribution process of the multi-brand channel between the FW17 collection and the FW20 collection, one obtained the predicted number of customers per collection until the year 2020. One then established the ratio between the total occupied area per activity for the SS17 collection and the total number of customers for the SS17 collection. By using the product of the values reached, it was then possible to predict the area which would be taken up between the FW17 and FW20 collections. These results are presented in table 2. Table 2. Prediction of the area occupied (m2) between the FW17 and FW20 collections.

Activities

Area (m2) SS17

FW17

SS18

FW18

SS19

FW19

SS20

FW20

Multi-brand Pick by line

1125

1185

1305

1400

1495

1589

1684

1779

Storage of finished product

750

790

870

934

997

1060

1123

1186

Raw material warehouse

422

445

490

526

561

596

632

667

Product expedition

200

211

232

249

266

283

300

317

Total

2497

2631

2897

3109

3319

3528

3739

3949

The analysis undertaken of the distribution process points to the following problems: • Electronic mail is used to send orders of the work to be executed by the multi-brand pick by line activity;

Tiago Silva et al. / Procedia Manufacturing 17 (2018) 655–662 Author name / Procedia Manufacturing 00 (2018) 000–000

660 6

• Display and manual printing of the BI (Business Intelligence) report to provide information of the customer’s order status; • Identification of boxes sent to customers is manually filled in; • Transporter labels and their respective invoices are manually printed. 4.3. Mapping of the proposed distribution process The performance assessment of the current distribution process pointed to a proposal of reversing the rack order for the multi-brand pick by line activity, which would also be divided into three zones: customers requiring the greatest number of completed boxes will be placed in proximity of the box closure zone and, consequently, near the storage area for finished products (see fig. 4). In the same manner, three supply zones were created at the beginning of the corridors for each type of customer. One also devised transport carts which were more ergonomic and adapted to corridor width. With regard to picking, the creation of a transport kanban significantly increased efficiency in the task of removing boxes, which impacts positively on article picking. An analysis of the product family allowed for the design of an organization model to be used in the storage of raw material. Consequently, the product types requiring more pallet movement were placed closer to the supply zone. The calculation of the distance covered validated an alteration to the layout of the storage area for finished products. Here, one reversed the order of arrangement of the pallets used for the storage of the finished product and moved this to a section which adjoins the pick by line activity. One additionally introduced two gravity carriers, which connect the box closure zone with the finished product storage area. It was thus possible to eliminate the task of transporting the boxes which had to be put away. (1) Raw material warehouse

(2) Multi-brand pick by line

(3) Storage of finished product

(4) Product expedition

Up

4

3 2 Up

Up

X2

1 25 m2

PASSAGEWAY FOR MACHINES

PASSAGEWAY FOR PEOPLE

STACKER

COMPUTER

AMERICAN PALLET

RACK

STRUCTURAL PILLAR

UNITS: METERS 8 700 m2

Fig. 4. Layout of the logistics warehouse at the fashion retail company (proposed distribution process).

In addition to the activities aimed at optimizing the layout and supply flow, the following improvements were also implemented: • The development of an application for product expedition, which will print invoices and transporter labels automatically; • The development of an application which will enable work orders to be sent automatically to the brand-line pick by line activities, as well as to the raw material storage section. Thus, it will no longer be necessary to send work orders by electronic mail; • The change in the packing list layout allows the operator to visually decipher all the information which is relevant for the tasks involved in the final distribution process; • The development of an application for the weighing of boxes.



Tiago Silva et al. / Procedia Manufacturing 17 (2018) 655–662 Author name / Procedia Manufacturing 00 (2018) 000–000

661 7

5. Results achieved The results one achieved can be seen in Table 3. By resorting to the proposed distribution process, one can increase the quantity separated by each distributor during an 8-hour shift by 294 articles. Furthermore, the minimum cycle time of 0,015 minutes/article can be reduced to 0,013 minutes/article, which will enable separating articles in the time available for the next seven collections. With occupied space now reduced on average by 47 m2 per collection, the average distance covered has decreased by 1 498 468 meters per collection. In addition, one was able to reduce the number of workers required for each collection by 5 on average. One also saw an increase of 535 boxes in the storage capacity of the finished product. Table 3– Results achieved in each distribution process. Metrics Current distribution process Proposed distribution process Space occupied (m2)* 3 310 3 264 Capacity of the PBL MM (customers)* 2 198 2 220 Capacity of the finished product warehouse 4 200 4 735 (boxes)* Capacity of the raw material warehouse (pallets)* 378 324 Productivity (articles/worker/8-hr shift)* 2 325 2 619 Cycle time (minutes/article)* 0,015 0,013 Distance covered (m)* 6 421 284 4 922 816 Human resources (workers)* 39 34 Investment (€)** 23 754,42 * Average values for the period between the fall/winter 2017 and fall/winter 2020 collections. ** Total investment value for the period between the fall/winter 2017 and fall/winter 2020 collections.

Difference 1,39% 0,99% 11,30% 14,29% 11,23% 13,33% 23,34% 12,82% -

Through the implementation of the improvements presented, which imply an initial investment of 21 010,54 €, and an average investment of 391,98 € for each collection (FW17 a FW20), one expects a total profit of 84 504,23 € by the end of the FW20 collection (see Figure 5).

Investment vs. Savings € 100.000

Value (€)

€ 80.000 € 60.000 € 40.000 € 20.000 €-€ 20.000

SS17

Investment/Saving -€ 12.931

FW17

SS18

FW18

SS19

FW19

SS20

FW20

-€ 2.215

€ 11.140

€ 24.526

€ 34.843

€ 53.616

€ 68.806

€ 84.504

Collections Fig. 5. Relation between investment and savings per collection.

6. Conclusions The pick by line activity of the multi-brand channel is known to determine the rate at which customers’ requests are met. By considering this fact and comparing takt time between the fall/winter 2017 and fall/winter 2020 collections, with a cycle time of 0,015 minutes/article, one was able to reach an important conclusion. Namely, if

662 8

Tiago Silva et al. / Procedia Manufacturing 17 (2018) 655–662 Author name / Procedia Manufacturing 00 (2018) 000–000

work was to proceed under the current conditions, it will be impossible to fill the customers’ orders within the time requested for the spring/summer 2019, spring/summer 2020 and fall/winter 2020 collections. By implementing the proposals described in this paper, one was able to increase the quantity separated by each distributor during an 8hour shift by 294 articles. The minimum cycle time of 0,015 minutes/article was also decreased to 0,013 minutes/article, which will make it possible to separate articles in time for the next seven collections. The following changes were also undertaken: an average reduction of 47 m2 in the space taken up for each collection, a decrease of 1 498 468 meters in the average distance covered for each collection, an average reduction of 5 workers per collection, as well as an increase of 535 boxes in the storage capacity of the finished product. In addition, and when compared to the current distribution process, there will be a greater capacity to ensure the distribution of the seven forthcoming collections to over 22 customers. The proposal presents an initial investment of 21 010,54 € for the spring/summer17 collection, and an average investment per collection of 391,98 € for the period between the fall/winter 2017 and fall/winter 2020 collections. This will result in a sum total of 23 754,42 €, which the company will be able to pay off in only six months, with a total accumulated profit of 84 504,23 € by the end of the fall/winter 2020 collection. Due to the work undertaken for this study, overall performance of the distribution process for the multi-brand channel was clearly enhanced. This points to the importance of optimizing warehouse operations at a fashion retail company and how this can positively impact on the improvement of the services rendered to the customer. This ultimate result is achieved through the minimization of costs, as well as by the reduction of delivery dates, and better quality in the service provided to the end customer. References [1] Rouwenhorst, B., Reuter, B., Stockrahm, V., Van Houtum, G. J., Mantel, R., & Zijm, W. H. (2000). Warehouse design and control: framework and literature review. European Journal of Operational Research (pp. 515-533). Elsevier Science B.V. [2] Baker, P., & Canessa, M. (2009). Warehouse design: a structured approach. European Journal of Operational Research (pp. 425-436). Elsevier B.V. [3] Accorsi, R., Manzini, R., & Maranesi, F. (2014). A decision-support system for the design and management of warehousing systems. Computers in Industry (pp. 175-186). Elsevier B.V. [4] Lu, W., Mcfarlane, D., Giannikas, V., & Zhang, Q. (2016). An algorithm for dynamic order-picking in warehouse operations. European Journal of Operational Research (pp. 107-122). Elsevier B.V. [5] Gu, J., Goetschalckx, M., & Mcginnis, l. (2010). Research on warehouse design and performance evaluation: A comprehensive review . European Journal of Operational Research (pp. 539-549). Elsevier B.V. [6] Atieh, A. M., Kaylani, H., Al-abdallat, Y., Qaderi, A., Ghoul, L., Jaradat, L., & Hdairis, I. (s.d.). Performance improvement of inventory management system processes by an automated warehouse management system. 48th CIRP Conference on MANUFACTURING SYSTEMS (pp. 568-572). Elsevier B.V. [7] Shiau, J.-Y., & Lee, M.-C. (2010). A warehouse management system with sequential picking for multi-container deliveries. Computers & Industrial Engineering. LVIII, pp. 382-392. Elsevier Ltd. [8] Dotoli, M., Epicoco, N., Falagario, M., & Costantino, N. (2015). An integrated approach for warehouse analysis and optimization - a case study. Computers in Industry. LXX, pp. 56-69. Elsevier B.V. [9] Rakesh, V., & Adil, G. (2015). Layout optimization of a three dimensional order picking warehouse. International Federation of Automatic Control - PapersOnLine. XLVIII, pp. 1155 - 1160. Elsevier Ltd. [10] Boysen, N., Briskorn, D., & Emde, S. (2017). Sequencing of picking orders in mobile rack warehouses. European Journal of Operational Research. CCLIX, pp. 293-307. Elsevier B.V. [11] Wanitwattanakosol, J., Attakomal, W., & Suriwan, T. (2015). Redesigning the inventory management with barcode-based two-bin system. 2nd International Materials, Industrial, and Manufacturing Engineering Conference (pp. 113-117). Elsevier B.V. [12] Bahr, W., & Lim, M. (2010). Maximising the RFID benefits at the tyre distribution centre. 12th International MITIP Conference The Modern Information Technology in the Innovation Processes of the Industrial Enterprises (pp. 202-209). Charlotte Høppner Gundelund Hans-Henrik Hvolby. [13] Caridade, R., Pereira, T., Ferreira, L., & Silva, F. (2017). Analysis and optimisation of a logistic warehouse in the automotive industry. Manufacturing Engineering Society International Conference 2017 (pp. 1096-1103). Elsevier B.V. [14] Koster, R., Le-duc, T., & Roodbergen, K. (2006). Design and control of warehouse order picking - a literature review. European Journal of Operational Research (pp. 481-501). Elsevier B.V. [15] Domingos, D., Respício, A., & Martinho, R. (2016). Using resource reliability in BPMN processes. Conference on Enterprise Information Systems/ International Conference on Project Management/ Conference on Health and Social Care Information Systems and Technologies/ ProjMAN/ HCist 2016 (pp. 1280-1288). Elsevier B.V. [16] Plenert, G. (2007). Reinventing lean: introducing lean management into the supply chain. (pp. 233-269). Elsevier Inc. [17] Gelmereanu, C., Morar, L., & Bogdan, S. (2014). Productivity and cycle time prediction using artificial neural network. Emerging Markets Queries in Finance and Business (pp. 1563-1569). Elsevier B.V. [18] Salonitisa, Y., Tsoutsanis, P., Litos, L., & Patsavelas, J. (2017). Improving the curing cycle time through the numerical modeling of air flow in industrial continuous convection ovens. The 50th CIRP Conference on Manufacturing Systems (pp. 499-504). Elsevier B.V. [19] Rohani, J., & Zahraee, S. (2015). Production line analysis via value stream mapping - a lean manufacturing process of color industry. 2nd International Materials, Industrial, and Manufacturing Engineering Conference (pp. 6-10). Elsevier B.V.