Zhizu Sun
University of Groningen
Faculty of Economics and Business MSc BA Operations & supply chains February 2009
Reduce cost and improve
speed
Reduce cost and improve speed
A research to improve the processes in Logwin Venlo
Author Zhizu Sun Student number 1741888
University University of Groningen (Rijksuniversiteit Groningen, RUG) Faculty Economics and Business
Program Master of Science Business Administration (MSc BA) Specialization Operations and Supply Chains
First Supervisor Dirk Pieter van Donk Second Supervisor Gera Welker
Company Logwin Air + Ocean The Netherlands B.V. Location Venlo, the Netherlands
Company Supervisor Loek Feller
Date February 17, 2009
Place Groningen, the Netherlands Version Final
Preface
In front of you lies the final report of my master study of Operations management & supply chains at the University of Groningen. I would like to take a few lines to thank those people who made this possible, even though these words can hardly express half of my gratitude.
Firstly, I would like to thank my supervisor at RUG, Dirk Pieter van Donk, for his constant concern, advice and feedback. The meetings during the research really inspired me and helped me to make a clear and systematic layout of the thesis. My gains from each meeting make the long commute between Venlo and Groningen worth it after all. The last advice from Ms. Welker, my second supervisor, really helped me make a good structure of my report. Hereby I express my gratitude for her insightful assessment.
Secondly, I would like to thank all my colleagues in Logwin Venlo, without whom this research can never be possible. Special thanks are to my company supervisor, Project manager Loek Feller, who contacted me in the first place and offered me the opportunity to carry out the project and provided me with constant support throughout the whole research period.
Last but definitely not least, I would like to express my gratitude to my parents and my family who supported me during my study in Holland. Their belief and generous care warmed me through the coldest winter days in this foreign land.
Executive Summary
Not satisfied with the current internal operations, the management of Logwin Venlo tries to improve the order throughput time and at the same time reduce operational cost as much as possible. Bearing this thought, the research question is formulated as:
How can Logwin Venlo reduce cost and at the same time decrease throughput time/increase speed?
In order to answer this research question, a research model is adapted from OM scan and several sub questions are formulated. This research uses both field and desk methods to help clarify the processes and the problems. As the core business of Logwin Venlo is warehousing, the research focuses on inbound and outbound operations in the warehouse. Literature regarding these areas is reviewed in order to provide a theoretical basis for analysis and recommendations. A conceptual model is developed to show the factors that influence cost and speed performances in Logwin Venlo after the review.
After checking the current performances and objectives, no clear-cut performance gaps are found for cost and speed. One conclusion is that the performance measurement system is not quite capable of monitoring all aspects of the performance objectives; another reason might be that performances of cost and speed are already quite satisfactory; or furthermore, the performance objectives were set too low to reflect any gaps in the operations.
Process diagnosis is carried out to see if there were any hidden problems in the processes from the conceptual model, which are not optimized in terms of cost and speed. Several problems are found in the three processes during the diagnosis: replenishment, mix pallet picking and order flow control in the warehouse.
These processes are discussed at length in the analysis phase. Two extra operational causes are found: communication and information system. When the problems and causes are known, improvement proposal is generated:
Improve communication with X
Reduce travel time for order picking
Travel time of mix pallet picking can be shortened using class-based storage. Although the trade-off might be more movements caused by class changes, cost saving is tested to be acceptable. Travel policy can be optimized by altering pallet building logic in the scanner. One possible way is to use volume of cartons instead of quantities to reduce the time of reshaping the pallets.
Shop floor control
Warehouse activities should be more controlled using documents issued in the office, especially for order sequencing. By doing so, waiting time for mix pallet picking and order related internal replenishment can be reduced.
Improve the performance measurement system
A more complete KIP system should be set up to monitor all aspects of performances in Logwin Venlo. This kind of measurement system will not only help find problems of past operations, but also in an early stage and make reactions immediately.
Table of Content
1. Introduction ... 10
1.1 The Company ... 10
1.2 Warehouse layout and storage type ... 13
1.3 The client ... 14
1.4 Main business process ... 15
2. Research design ... 17
2.1 Research motive ... 17
2.2 Research objective and question ... 18
2.3 Research model ... 18
2.4 Sub-questions ... 19
2.5 Boundaries ... 20
2.6 Methodology ... 21
2.7 Structure of the report ... 23
3. Literature review ... 24
3.1 Performance objectives ... 24
3.2 Factors that influence cost and speed in a warehouse ... 25
3.3 Performance measurement ... 29
3.4 Conceptual model ... 30
4. Objectives and performance ... 31
4.1 Speed ... 31
4.2 Cost ... 32
4.3 Relationship between cost and speed ... 36
5. Process description and diagnosis ... 38
5.1 Replenishment ... 38
5.2 Mix pallet picking ... 39
5.3 Planning and control mechanism ... 39
6. With-in process gap analysis ... 44
6.1 Replenishment ... 44
6.2 Mix pallet picking process ... 45
6.4 Extra-operational causes for performance gaps ... 49
7. Recommendations ... 52
7.1 Communication with X ... 52
7.2 Reduce travel time for order picking ... 52
7.3 Shop floor control ... 60
7.4 Improve the performance measurement system ... 61
Conclusion ... 62
References ... 65
Appendix I: Inbound process of Logwin Venlo ... 70
Appendix II: Outbound process of Logwin Venlo ... 71
Appendix III: Release schedule-One week in 2008 ... 72
Appendix IV: KPIs- On time arrival and error free shipments ... 74
Appendix V: 6 examples of routing policy ... 75
Appendix VI: Cost of two external warehouses ... 76
Appendix VII: Shipment waiting in Expedition ... 77
List of Tables
Table 2.1Research methodology and data collection ... 22
Table 3.1 Comparison among random, volume based and class based storage locations 27 Table 5.1 Replenishment types ... 39
Table 6.1 Time distribution during mix pallet picking ... 45
Table 6.2 Mix pallet-Picking list ... 46
Table 7.1 Average movements with relocation in every 3 working days ... 56
Table 7.2 Average movements with relocation in every 5 working days ... 56
Table 7.3 Average movements with relocation in every 11 working days ... 56
Table 7.4 Average SKUs in class A with relocation in different time laps ... 57
Table 7.5 Average SKUs in class B with relocation in different time laps ... 57
List of Figures
Figure 1.1 Group structure ... 12
Figure 1.2 Logwin Venlo-Company structure ... 12
Figure 1.3 Warehouse layout ... 13
Figure 1.4 Logistics network of Logwin Venlo ... 16
Figure 1.5 High level process mapping ... 16
Figure 2.1 Research model ... 18
Figure 3.1 Relationship between 5 performance objectives ... 25
Figure 3.2 Warehouse layout design and features (Bartholdi and Hackman 2005) ... 27
Figure 3.3 Conceptual model of cost and speed ... 30
Figure 4.1 On time departure (Jan. 2005 to Nov. 2008) ... 31
Figure 4.2 Warehouse man-hour per month (Jan.2008 to Oct.2008) ... 33
Figure 4.3 Warehouse labor productivity (Jan. 2008 to Oct. 2008) ... 33
Figure 4.4 Warehouse labor productivity-Inbound (Jan.2008 to Oct.2008) ... 34
Figure 4.5 Warehouse labor productivity-Outbound (Jan. 2008 to Oct. 2008) ... 34
Figure 4.6 Inventory level (Jan. 2007 to Oct. 2008) ... 35
Figure 4.7 Man-hour cost comparisom (2008) ... 36
Figure 5.1 Rack location ... 36
Figure 6.1 Time distribution during mix pallet picking ... 45
Figure 6.2 Routing of mix pallet picker with the picking list (table 6.2) ... 47
Figure 6.3 Shipment waiting in expedition area (Source: Appendix VII) ... 48
Figure 6.4 Number of pallets waiting in expedition area (Source: Appendix VII) ... 49
Figure 6.5 Extra-operational functional areas (Riezebo, 2003: 10) ... 49
Figure 7.1 Distribution of SKUs per picking (Sep. 2008) ... 53
Figure 7.2 4 types of layout in class based storage location implementation ... 54
Abbreviations
ADR goods Hazardous goods AAD Actor activity diagram
AS/RS Automated storage and retrieval system CBS Class-based storage
COI Cubic per order index
CAPS Computer aided picking system DC Distribution center
EC goods European goods ETD Earliest time departure FTL Full truck load
ICT Information and communication technology IC(IC/QC) Inventory control (Inventory and quality control) I/O point Input and output point
KPI Key performance indicator OM Scan Operation management scan
NEC Non-European goods/ Far East goods PO Purchasing order
PDCA Plan do check act SKU Stock keep unit
TF Transflow
1. Introduction
This chapter will present an overview of the company Logwin Venlo. Firstly, a glance will be taken to the history and present situation of the group at large. Later, focus will be drawn to Logwin Venlo, within which branch this research has been done and subsequent recommendations are aimed at. Later, attention will be drawn to the warehouse of Logwin Venlo because it is where most operations take place. Following that, X- a Sweden cosmetics company- the one and the only client of Logwin Venlo will be presented as well as the products the company provides worldwide. Last but not least, the main business process of Logwin Venlo will be described.
1.1 The Company 1.1.1 Logwin AG
Birkart Globistics was come into being in 1877 under the name of Birkart haulage in Aschaffenburg, Germany. In 1972, the establishment of the Hong Kong branch opened a new era for the development of the company. Ever since then, the company is conquering continent after continent. In 1989, the foundation of Sydney branch fulfilled the continental ambition, but another high speed expansion period began.
In 2002, Thiel Logistik AG took over Birkart Globistics Ltd (UK). As a logistics group, Thiel had several individual operating logistics companies. Birkart Globistics Ltd (UK) remains its own operation and services after the takeover.
In August 2002, Delton AG acquired the majority of Thiel Logistik’s share, and became the main shareholder ever since. A group wide re-branding campaign was launched on 1st. July, 2008, firstly in Germany, Austria, and Switzerland, and later worldwide. The new brand which unites the group strongly together is Logwin, and the fined and concentrated slogan “Your logistics” adds the brand with a feeling of customer intimacy.
Headquartered in Grevenmacher (Luxembourg), Logwin AG now provides a comprehensive range of logistics and service solutions for trade and industry. In 2007, the group generated sales of 2.0 billion Euros and currently employs approximately 8,500 staff in 44 countries. Logwin operates in all important markets worldwide and has over 400 locations across all continents.
special transportation activities in Central, Western and Eastern Europe) Logwin AG is one of the market leaders.
1.1.2 Logwin Air + Ocean The Netherlands B.V. (Logwin Venlo)
Logwin Air + Ocean The Netherlands B.V. is a branch of Logwin. It was founded in 1996, originally named Birkart Globistics B.V. under Birkart Globistics Ltd (UK). The intention of the project was to provide the client with good services of supply chain management in a strategically important location.
The Netherlands was the first choice because of its stable economical environment and unique advantages:
In the center of the pan European market State of the art terminal facilities
Advanced logistics network and infrastructures Government support and preferential policy
Venlo was chosen mainly because it’s geographical advantage. Venlo is the ideal location for Birkart to get access to X’s customers around the world because of the logistical resources of sea (Antwerp, Rotterdam) and air (Maastricht, Eindhoven, Schiphol) ports, rail and highways. After the establishment of the branch, several other improvement projects have been carried out. In 2004, barcode and scanner project were launched to manage the operations. Ever since then, the compatibility between ICT and barcode and interface of ICTs have roused countless discussions and followed up by solutions. Currently, the information system plays a vital role in Logwin Venlo’s operation.
During the rebranding campaign, Birkart Globistics B.V. transformed to Logwin Air + Ocean The Netherlands B.V.- the focus of this research and the place where it has been carried out. By using the abbreviation of Logwin Venlo, a distinction can be made from other Logwin branches.
Figure 1.1 Group structure
Staffed with around 45 full time and about 12 flexible temporary employees, Logwin Venlo manages one internal warehouse and a considerable amount of inventory in 5 other external warehouses.
The organizational structure of a company shows in what way tasks and responsibilities are divided into distinct groupings, and how the responsibility and coordination relationships between the groupings are defined (Slack et al, 2007). The organizational structure of Logwin Venlo is drawn below:
Figure 1.2 Logwin Venlo-Company structure
Managing director is in charge of Logwin Venlo and Logwin Piaseczno, because these two branches are both dedicated to X. IC/QC is the short form of Inventory control and quality control department. Project team is separated from operational departments, because it not only focuses on operations but also ICT and other projects. The full time equivalent employees in the warehouse are 27, while some temporary workers are hired through agent each day for mix pallet picking.
1.2 Warehouse layout and storage type
Figure 1.3 Warehouse layout
Warehousing is the core business of Logwin Venlo, the layout of the warehouse is depicted above. This internal warehouse has 10,000 m2 with rack locations of 14250 pallets. Hall 1 has 5 level racks, hall 2 and hall 3 have 6 levels and hall 1 has 3 levels. The ground level is for mix pallet picking. Checking area is the I/O point, because all outbound labels are distributed here and all mix pallets have to be checked here as well. The checked mix pallets will be moved to the automatic sealing machine. After sealing, they will be forwarded to expedition area. Recharge room is used for recharging electricity for equipments, such as forklift trucks. The repack room is where damaged cartons being repacked, damaged items cleaned and some requested-item-quality checked. Inbound area is for staging of inbound shipments. Also in here, the inbound shipments are checked and sometimes sealed (when they are transported in loose cartons). All docks can be used for loading and unloading; operators will choose the docks according to their experience, for example, replenishment shipment will be unloaded on the docks near pending area.
All inbound and outbound shipments are in pallet level and picking are down to case level. For inbound shipment, the strict rule of one item on one pallet is guaranteed. Also, one item can only have one pick location, unless it’s a wellness products, two pick locations could be possible.
1.3 The client
1.3.1 The company X
X is the only client of Logwin Venlo. X is founded in Sweden by the brothers Jonas and
Robert af Jochnick and their friend Bengt Hellsten in 1967. It is a cosmetics company that sells high-quality natural skincare and cosmetic products through an independent sales force, outside the traditional retail environment.
X Cosmetics is today one of the world's fastest-growing cosmetics companies. Present in 59
countries, X is one of the market leaders in over 30 countries. A sales force of 2.3 million independent Sales Consultants market a complete range of high quality skincare, fragrances and cosmetics. Logwin Venlo and Logwin Piaseczno are two branches of Logwin dedicated to X. All orders received or released in the two branches are by and for this one client. Geographically departed, the operations are closely connected, as well as the two independent but linked ICT systems.
As a cosmetics company, X enjoys a sales’ growth of 20% annually, despites the sloping economy. The growth imposes heavy loads on the supply chain of the company.
1.3.2 Products
As a cosmetics company, X not only sells cosmetic products but also wellness products and accessories. Due to different markets in different countries, one item should have different instructions thus result in a huge amount of SKUs. There are around 3000 SKUs on hand by Logwin Venlo, and they are still increasing on a daily basis.
Cosmetics
products which are not in the new catalog, a detailed products list can be found on the website of X for each market. Some of the cosmetics are sprays and has alcohol in them; they are marked as ADR goods (under the regulation of “The European Union adopts the United Nations Globally Harmonized System of Classification and labeling of chemicals, ADR 2009”). These goods should be stored where provisional equipments for fire are needed. The suppliers are mainly in Europe (Switzerland, Poland) and Asia (India, China, Russia)
Wellness products
Wellness products are newly introduced by X, which are mainly food supplements and vitamins. This kind of products need more care in warehouse operation. Temperature control and batch code traceability are required. Thus a current project has been done regarding wellness products traceability throughout the supply chain. This kind of products is in an introduction period, though the quantity is quite low now, it is expected to grow enormously in the near future. New challenges for Logwin will then be escalated.
Accessories
Accessories are mainly sourced in China. It can be cosmetics bags, necklace, ear rings and so. Because of the nature of these items, expire date is not applicable.
1.4 Main business process
X has a complex supply chain. It manages a lot of regional warehouses in the markets.
Figure 1.4 Logistics network of Logwin Venlo
The two Logwin Warsaw indicates that inventory transshipment is one of the important links in the supply chain of X.
By zooming in the lens to Logwin Venlo, the main business process is illustrated below:
Figure 1.5 High level process mapping
From this diagram, the process over view of Logwin Venlo is made clear. Orders are divided into two categories: inbound purchasing orders and outbound customer orders. Inbound process starts from pre-advise by suppliers and ends when the shipment is put away. When the shipment is on the racks, inventory control department takes over the role of managing it, making sure FEFO (first expire first out) and FIFO (first in first out) are applied, quality and expire date are monitored, replenishment is executed daily to facilitate outbound processes. Customer releases orders each day in TF. Outbound administrator receives an email regarding the orders and has to allocate the orders and print out documents and labels for warehouse operations. Warehouse activities are triggered by instructions from office. It could be a call from traffic department when the driver arrived, or outbound labels printed by outbound administrator. The customer decoupling point is the storage point in the warehouse, which decouples inbound and outbound operations.
Customer orders Purchase
orders
2. Research design
In this chapter, each step concerning this research is elaborated, serving as a guideline for doing this research. Firstly, why this research is carried out will be explained. Then the objectives of the research will be revealed. Furthermore, the main research question is posed and a research model will be deliberated to guide the pace of the research. Sub questions are derived from the model which will be answered in the following chapters. Next, boundary of the research is set and methodology used in this research will be covered. Last but not least, the structure of the report is clarified.
2.1 Research motive
Manufacturing in 5 countries and serving markets in 59 countries around the world, every node in the supply chain of X should be closely linked to each other. For a high profit margin product as cosmetics, one sale lost or one order not fulfilled not only means losing one customer but also means losing a potential seller due to the direct selling mode of X. So X is making all it could to maintain a high level of inventory in order to satisfy customer orders without delay. As the logistics service provider, Logwin Venlo has to manage the inventory of
X. Ever since September 2008, the purchasing orders are sky rocketing, resulting an under
2.2 Research objective and question
Research objective clarifies what relevance of the research is, in other words why the research is carried out. As mentioned in Chapter 1, X is a fast growing company with an average annual growth rate of 20%. In order to provide satisfactory services to the client, management of Logwin Venlo is continuously searching for improvements internally. In Slack et al. (2007), five performance objectives are deployed to measure organization performance. Among the five, cost and speed are of the most interest to the management. Based on the objectives, a research question was formulated:
How can Logwin Venlo reduce cost at the same time decrease throughput time/increase speed?
This research question contains two objectives: cost and speed. They are dependent variables which are affected by the independent variables, and between each other. By setting this research question, the aim and the direction of the research is clear.
2.3 Research model
Figure 2.1 Research model
objectives are divided into order winners and order qualifiers through analyzing customer wants and needs of certain goods or services, however this research is only internally focused to internal speed and cost. Another issue is that it's possible the performances are already meeting the objectives after the diagnosis, but still possible potential improvements and recommendations regarding optimizing the corresponding processes will be made. From these two perspectives, other performance objectives, say dependability, flexibility and quality will not be evaluated or discussed.
2.4 Sub-questions
In order to answer the main research question, several sub-questions are deducted. By answering these sub-questions, the research will be completed and the main question can be answered.
1. What factors influence cost and speed performances in Logwin Venlo?
As warehousing is the main business of Logwin Venlo, literature review regarding the two performance objectives, warehousing and planning and control will be done to answer this sub question in order to provide theoretical basis for process analysis and later recommendations. A conceptual model will be deliberated to point out the factors that influence cost and speed in Logwin Venlo.
2. What are the actual performances of the targeting objectives?
Slack et al. (2007) distinguishes five dimensions that constitute the performance of transformation processes: quality, speed, dependability, costs and flexibility. These dimensions are general, but the meaning of these performance dimensions varies per situation. As for this research, speed and cost are chosen to be evaluated, and they have specific meaning for this organization. This question will answer what do cost and speed, these two performance objectives stand for in Logwin Venlo.
As mentioned in the research design, due to the one client situation, order winner and order qualifier may not be very accurate if look at each order on a daily basis. In the long run, it is easier to find out which objective is order winner or qualifier, for example, if the client wants to establish a new DC in a new market, its choice could be based on one or a few performance perspectives of Logwin Venlo to decide whether to give the offer or not. In this case, order winner and qualifier may apply but only in a long term basis. Thus they will not be further discussed in this research.
Govil & Manish (2002) divides all business processes in a supply chain into 11 main activities. As a logistics service provider, Logwin Venlo mainly focuses on in/outbound logistics, inventory management and customer order processing. Firstly a global description of the inbound and outbound process will be given and then the processes in the conceptual model will be diagnosed as well as planning and control mechanism.
4. What are the relevant (sub) processes contributing to the gaps of the performance objectives?
Once the performances of the objectives are clear and the problems are known, the processes or sub processes which contribute to the performance gaps will be dug deep down. A proposed approach is to ‘staple yourself to the order’. The business object that is present through all process steps is the (customer) order. This analysis leads to the discovery of redundancies or incompatibilities within the processes and the planning and control applied to them.
5. What are the extra-operational causes for the performance gaps?
Riezebos et al. (2003) proposed several extra functional factors which will affect operational causes in the organization. This research will not access all the extra-operational causes but only those have effects on cost and speed performances.
6. What kind of potential improvements can be made? /If there's no gap, how can the process be optimized to reach higher objectives?
This is the last sub-question, which is directly linked to the main research question. Propositions regarding improvements will be provided if there are really some gaps between the performance and objectives.
It’s possible that before answering this sub question, no gap is found after diagnosis, or even the organization outperformed the objectives. In this case, a proposal for retuning the objectives will be given as well as recommendations for optimizing the process in order to reduce cost and increase speed further more.
2.5 Boundaries
The master thesis is part of the master’s degree program MSc BA Operations & Supply Chains and will thus be subject to the regulations of the University of Groningen. The period of this research is from 22. Sep. 2008 to 28. Feb. 2009, in total 5 months
The focus of this research is only on bulk operations, which means picking is down to carton level.
The capacity of internal warehouse can be increased but only through construction of another building beside the original facility, which means the capacity in terms of space can't be increased in a short time period, and of low cost.
The research only covers internal operations; transit times between suppliers to Logwin Venlo or Logwin Venlo to customers are not investigated.
As X determines inbound and outbound orders, there’s nothing much can be influenced by Logwin Venlo regarding order acceptance and order release.
All internal data concerning this research is in the scale of Jan. 2007 to Dec. 2008.
2.6 Methodology
2.6.1 Field research and secondary research
This project within Logwin Venlo is a combination of field research and desk research. In the book: In the Field: An Introduction to Field Research, Burgess & Robert (1984) pointed out that “Fields' that is circumscribed areas of study which have been the subject of social research”. Fields could be education or industrial settings. Desk research (sometimes known as secondary data or secondary research) involves gathering data that already exists either from internal sources of the client, publications of governmental and non-governmental institutions, free access data on the internet, in professional newspapers and magazines, in annual reports of companies and commercial databases.
2.6.2 Literature review
In order to set the theoretical base for process analysis, literature review will be carried out. Several literature regarding warehouse operations, planning and control mechanism will be discussed as they are the main business activities in Logwin Venlo. Literature regarding cost and speed, these two performance objectives and its measurement will also be discussed.
2.6.3 Observations & Interviews
processed and all the activities are coordinated. By following the orders and observing the actions of the employees’ toward the orders, all the information will be recorded for gap analysis.
2.6.4 Data gathering and analysis
In the second phase, data will be gathered in order to measure the performance. Logwin Venlo uses a WMS system which is called Transflow. It has two modules: WMS and Transport management. All order related information can be found in WMS and transportation related data in Transport management module. Consultation with the fist line supporting team will be made in order to minimize the chance to select the wrong data. All gathered data will be processed in the attempt of finding excessive cost and the wasted time which lengthened the order fulfillment cycle. Table 2.1 explicitly displays the steps, the methodology and means used and the data or information needed for each particular phase. The data/information column could be over or under listed, modification will be done during the research.
Phase Steps
Research
methodology Means Data/information
1 Literature review Desk research Literature review Theoretical background
2 Objectives and
performances Field research
Performance measurement Inquiry Process time
Speed Productivity
Cost Replenishment time Man-hour cost Cycle time
3 Diagnosis Field research Interview / Observation
Process description & diagnosis Planning and control mechanism
4 Gap analysis Field research/ Desk Inquiry Order pattern Picking routings Inventory transshipment Added value Replenishment cost
Theoretical throughput time
5 Extra-operational
causes Desk research Comparison Find extra-operational causes
6 Recommendations Desk research Turn back to literature
Recommendations for Improvement
2.7 Structure of the report
3. Literature review
This chapter combines literature review and conceptual analysis, and will provide theoretical base for process diagnosis, analysis and recommendations. Firstly, the selected performance objectives speed and cost will be discussed. Secondly, past researches of warehouse operations will be reviewed, especially on order picking, the most costly and time consuming activity in the warehouse. Next, planning and control mechanism will be shortly discussed. Later, literature regarding performance measurement will be examined for guiding later process analysis. Last but not least, the factors which affect cost and speed in warehousing will be summarized in a conceptual model. This chapter answers the first sub question.
3.1 Performance objectives
There are a lot of performances objectives used by different organizations and appear in different literature regarding 3PLs. In this research, Slack et al. (2007)’s five dimensions that constitute the performance of transformation processes are used and the focus is on internal cost and speed.
3.1.1 Speed
Speed refers to customer order delivery time, the throughput time of the complete order cycle; it is a shorthand way of saying ‘Speed of response’ (Slack et al, 2007). It means the time between an external or internal customer requesting a product or service, and them getting it. It is directly related with dependability, quality and flexibility and indirectly related with cost. There are internal and external effects of speed to an organization. Externally speed is important because it helps to respond quickly to customers. Again, this is usually viewed positively by customers who will be more likely to return with more business. Sometimes also it is possible to charge higher prices when service is fast. The postal service in most countries and most transportation and delivery services charge more for faster delivery, for example.
3.1.2 Cost
"If managed properly, high quality, high speed, high dependability and high flexibility can not only bring their own external rewards, they can also save the operation cost."
3.1.3 Relationship of cost and speed
Figure 3.1 Relationship between 5 performance objectives
This is a relationship diagram adopted from Slack et al. (2007). It explicitly specifies the relationship between different performance objectives. The dotted line between speed and cost represents an indirect affect between the two objects. There are two areas where speed reduces cost proposed by Slack et al. (2007): reducing inventories and reducing risks. The examples used in the book are from manufacturing but the same thing applies to service operations. But a shorter throughput time, will have effects in improving cost, productivity, and profitability (Blocher et al, 1999)
3.2 Factors that influence cost and speed in a warehouse
consider order picking as the highest priority area for productivity improvement (De Koster et al, 2007).
3.2.1 Order picking
There are several ways to classify order picking methods. De Koster, (2004) distinguished order picking by employing human or machines. Picker-to-parts is the most common picking method when employing human, while parts-to-pickers can often be seen in an automated storage and retrieval system (AS/RS). Robots picking and automated picking is only used in special cases such as valuable, small and delicate items. Order picking can also be classified as strict order, batching and zoning (Petersen et al, 2004).
The primary objective of order picking is to minimize travel distance, as travel time is normally the most time consuming component, 50% of this activity (Tompkins et al, 2003). Order picking methods cannot stand alone in a company’s strategy. Based on De Koster et al. (2007), the factors affecting order picking are classified into several categories: 1) Warehouse layout; 2) Storage assignment; 3) Routing methods. Because they are the components of order picking system, therefore, they are directly or indirectly speed and cost related, and will be discussed further.
3.2.1.1 Warehouse layout
Figure 3.2 Warehouse layout design and features (Bartholdi and Hackman 2005)
Hassan (2002) provides a framework for designing of warehouse layout. The article gives specific steps regarding warehouse layout design. From purpose of the warehouse to demand forecast and then the internal aisle, storage partition design. Among various steps, the author specified that class formation helps reduce picking time and distance.
3.2.1.2 Storage assignment
According to Petersen et al. (2004), three commonly used storage location assignment policies by most organizations are random, volume-based storage (VBS) and class-based storage (CBS).
Random Volume based Class based
Administration Easy Difficult Relatively easy
Information Little Much Medium
Time saving None Most efficient Efficient
Table 3.1 Comparison among random, volume based and class based storage locations
assigned warehouse locations within their respective storage class area. The storage class containing the highest volume SKUs is located nearest to the I/O point. Petersen et al. (2004) suggest using the CBS with 2-4 classes in practice, which is easier to implement than VBS but not require a complete list of items ranked by volume and less overhead administration. The article specified three important elements affecting the implementation of CBS: SKUs per picking list, storage location layout, item partition and class numbers.
3.2.1.3 Routing policy
The objective of routing policies is to sequence the items on the pick list to ensure a good route through the warehouse. Roodbergen (2001) distinguishes several heuristic methods for routing order pickers in single-block warehouse. The examples are given in Appendix V. Apparently optimal routing generates shortest travel distance but in practice, it may not be the best solution for warehouse optimization, for instance, congestion and different warehouse layout (De Koster et al, 2007). For information support system, a CAPS (computer aided picking system) or electronic paperless pick-to-light system can be utilized in practice in a picker-to-parts system. The CAPS automatically guides order pickers quickly to the pick locations, and shows the exact numbers of SKUs to be picked. The advantages of the CAPS include effectively improving the picking productivity by 50% or more, and reducing the picking task error (Jane & Laih, 2005). Routing policy is then inserted in the CAPS, and pickers just follow the instructions of the signal.
3.2.2 Replenishment
Internal replenishment is different from external replenishment. It plays a key role in the warehouse as it ensures that there is enough stock in the pick area to cover the expected demand. The replenishment in this context means intra-warehouse replenishment from a reserved or bulk area that is optimized for storage to a small and compact pick area optimized for unit picks (Aditya, 2008). Therefore, it is often speed related. This article classifies items according to demand and pick frequency and then rank them into different classes: fast, medium and slow moving items. And it provides different replenishment scheme for these classes. This article is originated from Oracle’s WMS; hence the data is all available through the system.
3.2.3 Planning and control mechanism
Schönsleben (2007, p476) classifies planning and control in complex logistics system as 8 different perspectives. As the author himself noted, many of these techniques have their origins in the MRP II and ERP concepts. However, they also apply to process industry as well as to non-repetitive production and non-repetitive procurement. From the 8 perspectives, it’s obvious that they have everything to do with speed, such as order release and control or time management and scheduling. Its relationship with cost is not that explicit, but from the perspective of inventory control, it is cost related.
Slack et al. (2007) defines planning and control more straightforwardly. “Planning is the theoretical end of the activity, while control is the more applied end. Planning involves deciding what to do and when to do it. Control involves making sure that plans are actually taking place in practice and responding when things do not according to plan. Planning looks at activities sometime in the future. Control looks at activities that are happening now.” He proposed four questions to be answered during planning and control.
Scheduling – when should we do things? Loading – how much should we do?
Sequencing – in what order should we do things?
Monitoring and control – are things going to plan? If not, what should we do about it?
3.3 Performance measurement
point to discuss the performance of outsourcing third party logistics service. They suggest using KPIs to measure the performances of logistics service providers. Their survey found that consumer goods companies perceive transport to be the best and information systems the worst suited for outsourcing. A large number of practices are in place for managing storage. Companies looking to outsource warehousing often face the lack of suitable facilities in terms of capacity and specialist requirements.
3.4 Conceptual model
Figure 3.3 Conceptual model of cost and speed
From the literature review above, a conceptual model is depicted. All three factors will influence both speed and cost performances.
Order picking process is strongly cost related because it occupies the majority of warehouse expense. Also, it determines the time when the order is ready to be shipped, therefore, it is speed related.
As replenishment process can be divided into external replenishment and internal replenishment. They influence cost and speed differently. External replenishment mainly affects cost, because it involves inventory and transportation management. Internal replenishment mainly affects speed, as its purpose is to make stock available in the picking location.
Planning and control is speed directly related, because it decides the order flow and prioritization of the orders. On the other hand, it involves inventory management, capacity management; therefore, it also influences cost.
4. Objectives and performance
This chapter will answer the second sub question, the current performance of the two objectives in Logwin Venlo. If there were any performance gaps, they will be identified in this chapter. Currently, there are three KPIs used in the organization: on time departure (Figure 4.1), on time arrival and error free shipment (Appendix IV). As on time departure is speed directly related, it will be used to evaluate performance of speed. While cost is not reflected in the three KPIs, it will be indirectly evaluated using the data from ICT and other management system.
4.1 Speed
For Logwin Venlo, as mentioned in boundaries, only internal speed/throughput time will be investigated. On time departure measures the promised order cycle time and the actual throughput time of the organization. Although on time arrival is indirectly related to internal throughput time, it is a measurement for the performance of the organizations of how they keep delivery promises. Hill (1989) describes delivery reliability as the ability of the firm to deliver on or before the promised scheduled due date. Thus on time departure is speed oriented and has everything to do with internal operation time, waiting time and the promised order cycle.
By looking at the figure 4.1, on time departure has improved tremendously during the past few years. The goal is set as 95% of all the orders, which means if one order is not on time dispatched, it is counted as 0. The data in 2008 showed an average of 99.75%. Except in Jan. May and July, it remains at 100% throughout the year- all on time.
Figure 4.1 On time departure (Jan. 2005 to Nov. 2008) 86% 88% 90% 92% 94% 96% 98% 100%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
This performance reflects that speed is already above the performance, which in turn implies that speed under current promised order cycle is satisfactory. But it could also be the case that the order cycle promised is too long and the process is not optimized in terms of efficiency. Due to the insufficient KPIs for measuring these aspects of speed, a closer attention will be paid to the process diagnosis in order to find the rational of the objective, the hidden gaps and potential improvements.
4.2 Cost
Cost is itself internally focused, thus internal financial data is the source of tracing cost throughout the operations of the organization. Usually, cost can be identified as: material costs, personnel costs, inventory costs, costs of using machines and facilities, coordination costs and so on. Due to the nature of inventory- customer owned, Logwin Venlo cannot change inventory level or manage inventory by altering inbound or outbound orders, thus cost of inventory cannot be influenced unless customer wants to lower them itself. Given this insight, the attention has been drawn to cost of warehouse operations, which is the major part of annual expenditures.
4.2.1 Cost of internal warehouse
Figure 4.2 Warehouse man-hour per month (Jan.2008 to Oct.2008)
It’s obvious that mix pallet picking consumes the largest amount of warehouse man-hour, because it is done by manual work. From this figure only, one cannot tell the relationship between the output (pallet volume) and the input (man-hour). As suggested by McGinnis (2003), one way to measure warehouse labor productivity is items per man-hour. Due to the bulk operation in the warehouse, pallet per man-hour is more appropriate to measure productivity. A productivity diagram is drawn where pallet volumes are divided by inbound and outbound operations respectively:
Figure 4.3 Warehouse labor productivity (Jan. 2008 to Oct. 2008)
When looking at the diagram, inbound activities, namely unloading, inbound check, carton sealing and put away followed the tide of inbound shipment, meaning as quantity increases, productivity increases as well. It is quite reasonable and it shows a stable performance throughout the whole year. While looking at the outbound productivity, all activities seem to decrease along the time line. There could be gaps in these activities.
0 500 1000 1500 2000 1 2 3 4 5 6 7 8 9 10 11 H ou rs Months
Warehouse man-hour per month
Two Volume-productivity diagrams are drawn to show the relationship between inbound and outbound volume with productivity of warehouse activities. The inbound productivity (Figure4.4) indicates that linear relationship is quite observable for inbound unloading, inbound check, carton sealing and put away.
Figure 4.4 Warehouse labor productivity-Inbound (Jan.2008 to Oct.2008)
Figure 4.5 Warehouse labor productivity-Outbound (Jan. 2008 to Oct. 2008)
Productivities of these activities increase with the increase of inbound volume. It shows that inbound capacity is not very flexible; when the workload is low the productivity is low while the workload increases, as well as the productivity. One conclusion is that during the off-peak period, over capacity may exist.
For outbound activities (Figure 4.5), especially for loading and picking full pallets, there’s no significant relationship between productivity and volume. The distribution seems like a “U”
shape. Only from this diagram, it gives no further explanation. A closer look will be paid during process diagnosis to see if there were any gaps in these activities.
4.2.2 Cost of external warehouses
As mentioned previously, 4 external warehouses are used for storing excess inventory. They are Van den Anker (Eindhoven and Roosendaal), Rhenus (Venlo), TWO (Maastricht) and EFS (Reuver). The inventory in each of the warehouse can be clearly seen from the inventory diagram below:
Figure 4.6 Inventory level (Jan. 2007 to Oct. 2008)
The stock level (Figure 4.6) has an increasing trend over the last two years. From the second half of year 2007, internal warehouse storage utilization is always higher than 90%, which means warehouse is running at its near full capacity and inbound shipment is usually not able to be unloaded due to lack of storage location and therefore be unloaded in external warehouses. When one carton or several pallets are ordered but not stored in the internal warehouse, inventory should be transferred from external warehouses.
Costs for external warehouses also vary from each other depending on different services. In future, only the external warehouse for storing ADR goods will be kept, thus the cost of man-hour for these warehouses and internal warehouse are compared.
0 5,000 10,000 15,000 20,000 25,000 2007.1 2007.5 2007.9 2008.1 2008.5 2008.9 Pa lle t Month
Total stock on hand
Figure 4.7 Man-hour cost comparisom (2008)
The figure above shows labor cost per hour only, not including other services and their relevant costs (Appendix VI). Thus for Logwin Venlo, cost not only means internal management and operation cost but also external cost for storage and inventory transfer, and external cost is higher than internal cost. In this sense, maxmizing the utility in the internal warehouse is rational.
4.2.3 Cost of inventory transshipment
From the supply chain of Logwin Venlo (Figure 1.4), it’s quite obvious to find that Logwin Piaseczno (WDP) at both ends. This is because inventory transshipment consists of a huge part of the inbound and outbound shipment. There’s no literature explicitly points out whether transshipments can save time and cost to what level, but Nonås & Jörnsten (2007) reckon that transshipment is mainly a result of better integration of the information systems.
By checking the data of inbound and outbound shipments for 2008, inventory transfer between Logwin Venlo and Logwin Piaseczno (WDP) is very frequent. While for Logwin Venlo and Piaseczno, the data showed that 25% of the outbound orders are shipped to WDP (20.5%), and 15% of the inbound orders are incoming from WDP. Due to no autonomy over inventory level, these decisions are totally made by X. There could also be gaps because of this big amount of inventory transfer. But as the decision is not made by Logwin Venlo, this kind of cost cannot be influenced much by improving internal operations.
4.3 Relationship between cost and speed
When consider the relationship between cost and speed in Logwin Venlo, it’s not quite easy to find it positive or negative directly. A faster internal speed will not necessarily mean a reduction of inventory, because of the fact that client manages inventory level. But higher
0 10 20 30 40 50
Van de Anker Two Internal
Man-hour cost
speed means higher response to customer orders and customer satisfaction. A shorter throughput time means a higher productivity both in office and warehouse activities and in return internal operational cost can be reduced. The approaches to achieve this faster throughput may also involve investment. Therefore, some of the recommendations will have to be tested regarding their potential saving and the trade-off effects between these two performance objectives. As this research is operations related, the relationship between these two performance objectives can be interesting for future studies.
Conclusion
5. Process description and diagnosis
To gain insight into the operations of Logwin Venlo, detailed process mapping is carried out. Firstly, the inbound and outbound process maps are drawn by AAD (Appendix I, II). Alongside the process mapping, the processes which affect cost and speed in the conceptual model will be diagnosed. This chapter will also answer the third sub-question.
5.1 Replenishment
5.1.1 Internal replenishment
Internal replenishment should be done every day independent of orders. Internal replenishment list is generated by IC administrator through replenishment module in TF in the morning and should be done within the day in the warehouse. It is aimed to move the bulk pallets from storage location to pick location before order picking, so that the workload for order related replenishment can be decreased and utilization of equipments and labor can be increased. The racking system in the warehouse is divided into different levels (H2 and H3, 6 levels; H15 levels and H4 3 levels), from A to F in TF respectively as illustrated on the left. Although some racking rows have AB as pick locations (B is blocked in TF), mix pallet pickers can only pick cartons from A level, B level in this case is made for internal replenishment and blocked in TF. In the system, there are two types of pick locations, single and double. The quantity for internal replenishment is different accordingly. To trigger single location replenishment, the item must be lower than 25% of its original quantity, while for 50% for double location.
5.1.2 Order related replenishment
Order related replenishment could be further divided into two types: external order related replenishment and internal order related replenishment. Internal order related replenishment is similar to internal replenishment, the only difference is one or more cartons in the pallet have been allocated to a specific order due to the insufficient of items in current pick location.
Storage location (Full pallets store in bulk)
F E D C B Pick location A
External replenishment is to transfer inventory from external warehouse. There are 3 different destinations for these pallets:
1. Go to pick location, for one or more cartons in the pallet have been ordered 2. Go to bulk location, for the pallet is to replenish bulk inventory
3. Stay in the pending area, for the full pallet is ordered and will go into the shipment.
Table 5.1 Replenishment types
Table 5.1 specifies all replenishment types and destinations for the replenished pallets. One problem is that order related replenishment often starts at the same time of mix pallet picking, so that unfinished mix pallets can often be found in checking area. Another problem is that in order to save transport cost, a full truck load is made during external replenishment, but what kind of items should be moved back from external warehouses is always an ad hoc decision.
5.2 Mix pallet picking
From the man-hour data (Figure 4.2), mix pallet picking is of least productivity; therefore, the process of mix pallet picking will be diagnosed here.Normally, checkers collect outbound labels, replenishment list and control sheet from outbound administrator. There are two different pallet labels: full pallet labels and mix pallet labels. They are distinguished by F (full pallet) and M (mix pallet) before pallet number. For mix pallet picking, picker will get one label for picking and the other one is for checkers to easily find the pallet which is not checked yet. When mix pallet picker gets the label, he/she has to check whether the market requires the fumigated pallet or not. The requirement is on a list in the checking area. After picking the right pallet, picker will scan the outbound label to start picking cartons. Picking process is quite simple: Scan, pick and confirm. Because of the low productivity (Figure 4.3), a field investigation is carried out to see the routing of the pickers and time distribution for mix pallet picking in Chapter 6.
5.3 Planning and control mechanism
Planning and control have many implications in different organizations, Slack et al. (2007) summarized four common procedures that most organizations encountered: Scheduling, loading, sequencing and monitoring.
Replenishment External Pending Bulk Pick
Internal x x from to
Order related internal x x from to
5.3.1Scheduling
Scheduling can be simply interpreted as “when should we do things?” In Logwin Venlo, pre-advise serves as inbound schedule for actual inbound shipment because it is usually sent days before shipment arrival. While for outbound, some agreements regarding scheduling had been made with the client when this branch established.
1. There are no backorders. X can only order the item and its quantity in the system, if they order more quantity than system registered, the quantity will be reduced in accordance to the physical inventory in stock.
2. A release schedule is used. It serves as an indication for how many and which markets will be released on each week day before what time. (Appendix III)
3. Rush orders. Rush orders are usually low in quantity but require fast shipment. X is free to place rush orders.
Except for rush orders, these agreements set an order pattern for order release by X: it usually releases the orders from several scheduled markets on one day preventing the situation of releasing orders for all markets from happening and it cannot order more than Logwin Venlo has both in the information system and in warehouse. If the order is released later than in the agreement, it will be treated as next day’s release. For most of the countries, the order cycle is defined as A-C (D) cycle, each letter represents one day.
Day A: Outbound receives orders from X, inventory control places order related external replenishment, and traffic books trucks.
Day B (B-C): Replenishment of orders from external warehouse; Warehouse pick, check and pack orders, outbound invoices and customs prepares documents
Day C (D): Warehouse loading
For some other countries, which require containers and sea transport, this cycle is longer, due to container liner schedules.
Orders from most of the markets should be released before 15:00. Release schedule is assumed to give forecast for order release each day, providing scheduling for resources in warehouse. But the data from 2008 shows a low accuracy of fulfilling these scheduled orders. The average accuracy is 47% (Appendix III).
5.3.2 Loading
dependent on the orders released from the client, which leaves the capacity of warehouse to a main concern. There are three constraints in warehouse for loading capacity: space, equipment and labor.
5.3.2.1 Space
The pallet locations as mentioned above in internal warehouse are 14250 in total, which is not enough for storing all the inventory. Logwin Venlo classifies items into EC, None-EC, ADR2, 3, 4; naturally, warehouse pick locations are classified into these areas which are illustrated in Figure 1.3. This kind of classification is not according to the nature of items. Because of the ADR regulations, the storage and pick area are relatively fixed, while the other items can be stored in any location in the warehouse. This kind of random storage policy may result in long travel time for order picking (Petersen et al, 2004); whether it will affect the order picking activity in Logwin Venlo will be analyzed in the following chapter.
5.3.2.2 Labor
Labor capacity in terms of mix pallet pickers is very flexible, because currently, mix pallet pickers are hired through agency. Every day around 4 o’clock, when outbound collects all the released orders, the demand for mix pallets are translated to determined demand of man-hour. Warehouse supervisor contact the agency to make sure the exact number of pickers will show up the next day. This method gives high flexibility for mix pallet picking, but when considering forecasting demand for a longer period of time or for other activities in the warehouse, it gives no indication at all.
March, there’s only one driver for picking full pallet. Another full pallet picker is added during April. In May, it increased to 3 pickers, thus the productivity decreased from March to May and remains stable afterwards.
5.3.2.3 Equipment
Equipment such as counter balance, reach truck, low lift electric pallet truck and scanners are used fully and set at a maximum level that can satisfy each operator with their specific tasks. All equipments can be used by different users. As for scanner, logon by scanning identity badge is required before using it for a specific task. From observation and interview in the warehouse, the situation of waiting for equipment is rare, thus this is not a constraint of warehouse resources.
5.3.3 Sequencing
Sequencing is the question of “In what order should we do things?” By giving priorities to different activity or orders, the process flow can be optimized. In Logwin Venlo, rush orders are recognized as express and air shipment, and they are given priority over other markets, because they have the nearest ETD (Estimated time departure). Beside rush orders, due to the promised order cycle for normal orders, A-C markets will be given priority in office operation during outbound process, which means outbound administrator usually print out the label of markets which has the shortest ETD. So, A-C markets are usually picked first and then A-D markets. But when the two markets are all released and printed out, outbound administrator will start the other markets, which usually require sea transport. Due to the liner schedule, these shipments will usually depart in the next week. When the outbound labels are printed, warehouse checkers will get the labels and distribute them to pickers. Although checkers know that markets with the shortest ETD should be picked first, but in practice, it is not always sticking to. Therefore, order flow in the warehouse is not always controlled in accordance to the sequences pre-set in the office.
5.3.4 Monitoring and control
mentioned above, there are only three KPIs available; they measures dependability, quality and external speed. Although all the data is available in TF, decision makers can’t easily retrieve it from the system, let alone measuring performances or making decisions regarding improvements.
Conclusion
6. With-in process gap analysis
From the diagnosis in Chapter 5, some problems in the process or in planning and control are found. This chapter will go into details to explore the reasons behind the problems. Extra-operational causes will also be shortly analyzed. This chapter is going to answer the fourth and fifth sub-questions.
6.1 Replenishment
6.1.1 Order related internal replenishment
Due to the standard dimension of the rack location, bulk pallets cannot fit into pick location when it is still occupied by a certain amount of cartons from previous pallet. This situation often occurs during order related replenishment because the quantity of items in pick location is not enough for mix pallet picking. As mentioned above, internal replenishment has a minimum item quantity to trigger internal replenishment while for order related replenishment there’s no such rule. It’s quite imaginable, otherwise, the quantity is not enough for mix pallet picking, and the replenishment loses its meaning. But this also causes a problem of double replenishment. When the items in pick location are not enough for mix pallet picking, TF automatically generates order related replenishment list with no regard to whether it could fit into the location or not. In this circumstance, replenishment operator, who has no knowledge of the still occupied location, has to move the pallet back to its original location and wait until mix pallet picker has finished the items on the pick location and required the replenishment for mix pallet picking.
6.1.2 Order related external replenishment
As mentioned in previous chapters, inventory stored in the external warehouse will be reduced to ADR goods only, due to the regulations in ADR goods in warehouse. Therefore external replenishment is mainly ADR goods. A fact is ADR goods only contribute a small portion, 6% of the total SKUs (2008). While the replenishment rate for outbound shipment is 21% for year 2008, which means 21% of the total outbound pallets are transferred from external warehouses. From the stock level data (Oct. 2008), 52% items stored in external warehouses are ADR goods, the rest are regular items.
may not be ordered in a long time period while other external stock may be ordered tomorrow and needs to be transferred again the next day. The storage cost for external warehouse is only 0.2-0.3 euro per pallet, while the ordering cost per pallet is around 18-24 euro. In this sense, the less transfer the lower cost it incurs. This rule applies to speed as well: the less external replenishment the faster an order can be fulfilled. The reason is if an order needs replenishment, it will spend one day for inventory transfer and replenishment, full pallet and/or mix pallet picking have to wait until replenishment is done. The full utilization of internal warehouse in turn forces incoming shipments to be unloaded into external warehouses, causing more future inventory transfer and replenishment cost.
Considering the limited rack location of internal warehouse, the use of external warehouses is inevitable. But there’s no clear information for inventory control department to support decision making when decide which item to store where.
6.2 Mix pallet picking process 6.2.1 Mix pallet picking
According to Bartholdi & Hackman (2005) “travel time is waste. It costs labor hours but does not add value”. It is, therefore, a first candidate for improvement. In Logwin Venlo, mix pallet picking is done manually. By following mix pallet picker in the warehouse (30 mix pallets), the average time distribution within this process is listed below:
Activity Scan labels
Pick a
pallet Travel
Wait for
replenishment Pick others
Seconds 603 723 8714 3611 6001 1637
Percentage 2.83% 3.42% 40.92% 16.95% 28.20% 7.69%
Table 6.1 Time distribution during mix pallet picking
A diagram is drawn to better illustrate the time distribution:
Figure 6.1 Time distribution during mix pallet picking 0.00% 20.00% 40.00% 60.00% 80.00% 100.00%
Travel Pick Waiting time others Pick a pallet Scan labels Mix pallet picking
This graph shows that travel time is around 40% of the total process time, pick time is second and waiting for replenishment comes third. The “other” includes throwing away the plastic film during replenishment, fetch the outbound label, count cartons and sort the pallet.
The current warehouse layout only allows pickers to adopt return policy during picking, and the routing is dependent on the scanner module of TF. This logic follows several rules:
1. Aisle search: firstly, TF will search each aisle in sequence to see if it contains the item for mix pallets
2. Within aisle quantity sequence: if several items should be picked in one aisle, it makes the picking for these items in a quantity-wise descending order down to 5 cartons
3. When the rest of the carton is less than 5 quantity each, a route optimization is made 4. Pallet height limit: the pallet should not be built higher than the pre-selected height, thus
dimension of each carton is considered within a tolerable level.
5. No same item on two mix pallets for one market: if adding one item will make the height exceed the limit, the quantity of the item will not be split into two mix pallets, a new pallet will be used for this item
6. ADR and bonded goods: ADR and bonded goods should be separately picked and minimize the chance of mixing with EC goods.
These rules seem to be reasonable on paper, but when applying them into warehouse pallet picking, sometimes, it is still not optimal. Take one mix pallet for example, its picking list is specified in table 6.2 below:
Sequence Hall Row Location Quantity
1 H1 G05 030 27 2 H1 G05 013 20 3 H1 G05 059 13 4 H1 G05 053 8 5 H1 G05 048 6 6 H1 G05 014 4 7 H1 G05 033 1
Table 6.2 Mix pallet-Picking list
Figure 6.2 Routing of mix pallet picker with the picking list (table 6.2)
This list is generated strictly following the rules set in the system, but as can be clearly illustrated, the picker has made some extra walks following the instruction of the scanner which can be avoided if the route is sequentially optimized. The attempt to select the biggest amount to pick first is to make a more stable pallet during picking, but it was based on cartons quantity without consideration of carton dimensions. Different cartons have different dimensions, for instance, items like lipsticks are packed in a more flat carton while creams or gels are packed in bigger boxes. The first item from the picking list above has a quantity of 27, but actually they are packed with inner boxes inside a big one, 1 big box has 12 inner boxes. Therefore 27 make only two big box and 3 small boxes while the second item has 20 big boxes, which is much more than the first item space-wise. The same thing happens for the third item, the carton is smaller than the rest of the items but the quantity is larger, and the last item though only needs one carton but appeared to be the biggest of all. One foreseeable result is that in order to make the pallet stable, picker usually has to leave the small cartons which cannot fit, on top of the pallet and reshape it after finishing all the items.
By checking the quantity in one month’s mix pallet (09.2008), nearly 47% of the pallet routing has this return problem, from one aisle to another and back to this aisle again or from one location and later return to the location next by. This finding confirms the logic in TF which will result in unnecessary travel as well as extra work for pickers to stabilize the pallet.
6.2.2 Waiting time during mix pallet picking
