Chapter 7.1 Material Handling and Storage Systems William Wrennall The Leawood Group Ltd., Leawood, Kansas Herbert R. Tuttle University of Kansas, Lawrence, Kansas 1.1 INTRODUCTION Material handling and storage systems planning and design are subsets of facilities planning and design. Material ¯ow has both internal and external eects on a site. There are in¯uences for the site plan and the operations space plan. Conversely, the material handling system impacts the facility plans, as illu- stratedinFig.1. In the facilities design process the material move- ment determines the ¯ow paths. The material move- ment origins and destinations are layout locations. The storage locations and steps are eects of the opera- tions strategy and thus the organization structure. A lean manufacturing system may have material delivery direct to point of use replenished daily and a pull sys- tem cellular manufacturing process that produces to order with a TAKT* time of 5 min. Such a system could have inventory turns of 300 per year. A more traditional system would have a receiving inspection hold area, a raw material/purchased parts warehouse, a single shift functional layout batch manufacturing system, inspection and test with 90% yield, a separate packing department, and a policy of one month's ®n- ished goods inventory. The space plans for the tradi- tional system should be very dierent from the lean approach and so should the material handling and sto- rage plans and systems. A ``pull '' system also indicates unnecessary material in the system. If it does not pull it should not be there. Material handling determines the capacity of a man- ufacturing plant. From the receiving dock to the ship- ping platform the material ¯ow routes are the circulation system. Flow restrictions can act as capa- city limiters. The material handling and storage plan determines handling and storage methods, unit loads and containerization to support the operations and business strategy. The product volume plotÐthe plot of volume/quan- tities of materials by product typically shows a nega- tive exponential distribution, the important few and the trivial many Pareto distribution. The plot can be overlaid with the most suitable production mode as illustrated in the product volume (PV)/mode curve, Fig.2. We have suggested the following modes: 1. Project 2. Cellular 3. Linked cellular 4. Line 5. Continuous. 607 * TAKT time is the rate at which your customer requires product. Copyright © 2000 Marcel Dekker, Inc. We ®nd these classi®cations more meaningful than ®xed location, job shop, functional, mass production line, and process ¯ow. In the manufacture of discrete solid parts their transportability is improved by placing them in con- tainers. This contained regular shape becomes the unit load and the material handling method is matched to the load. As material ¯ow volumes increase, the char- acteristics trend to those of continuous ¯ow, i.e., from solid discrete parts to bulk (¯owable) solids, liquids, and gases. Solids are handled as pieces or contained in baskets, cartons etc. Ware [1] describes how solids can also be conveyed by vibratory machines. Liquids and gases are always contained and they conform to the shape of the container. The container may also be the channel of material transfer, such as a pipeline or duct, particularly for ¯owable bulk solids. Bulk solids can also be transported along pipes or ducts with the aid of a suspension agent, such as a gas or liquid. In just-in-time (JIT)/lean manufacturing the aim of batch sizes of one is to emulate continuous liquid or gaseous molecular ¯ow characteristics to achieve simpli®cation. Designs for material handling in liquids, gases, and bulk solids are an integral part of the process. Examples are chemical and food processes. In an oil 608 Wrennall and Tuttle Figure 1 Material ¯ow patterns. Copyright © 2000 Marcel Dekker, Inc. re®nery the input of raw materials to shipment of pro- ducts is often totally integrated. In process design it is always desirable to move up the PV/mode curve. Process eciency increases from project mode to continuous ¯ow. This economy was achieved originally by increasing volumes and thus increasing inventories. More recently, economies of operations with low volumes have been achieved by a change of focus/paradigm through rapid added value to materials yielding increased return on assets employed. The assumption had previously been made that the economies came from volume only. However, material handling unit volumes and storage require- ments have shrunk with the use of: Family manufacturing Cellular operations Product focus Rapid setups Pull systems Batch sizes of one Make to customer order Mixed product less than full load shipments Half machines Lean operations. Material movement dominates the design of many facilities. Conversely, the layout design sets the loca- tions and origins of each material move. Movement adds cost, time, and complexity to manufacturing and distribution. It adds no ultimate or realized value until the ®nal moveÐdelivery to the customer. The priority therefore is to reduce material movement. The minimizing of material movement requires an eective layout based on a sound manufacturing strat- egy. The anities and focus approaches which can reduce both the amount and complexity of handling are powerful tools. They are described in Wrennall [2]. Layout aects material ¯ow in three ways. First, the space planning units (SPUs) de®nitions set a pattern foroverallmaterial¯ows.Figure1,givenearlier,illus- trates how production mode impacts both the intensity and complexity of ¯ows. Second, the arrangement of SPUs in the layout can increase or decrease particular route distances. Third, the arrangement of SPUs sets a large-scale ¯ow pattern (or nonpattern in some cases). Since layout design and material handling are inter- dependent, so is a discussion on the design of either or both of them. 1.2 MATERIAL FLOW ANALYSIS Material ¯ow analysis (MFA) examines the movement of materials over time. It helps develop anities for the layout and evaluation of layout options, and is basic to the design of material handling systems. Unlike other reasons for anities, material ¯ow is tangible and mea- surable. The use of quantitative methods adds rigor to the facility layout planning process. After rigorous analysis and simpli®cation, the remaining and neces- sary material moves are economical and timely. Ultimately, all assembly processes are material handling. This discussion limits the handling of mate- rials to and from a site and from place to place within the site. Material handling at the workplace and asso- ciated automation is discussed elsewhere. The objectives of material ¯ow analysis (MFA) are to compute anities based on material ¯ow, evaluate layout options, and assist handling system design. A macrolayout of 30 SPUs has 435 possible material ¯ow routes. In addition, most facilities have many materi- als, processes, handling methods, and multiple move- mentpathswith¯owsinbothdirections.Figure3 illustrates some of the possible material variety. Handling equipment, containers, route structures, and control methods all present additional variety. Therefore, analysis and the subsequent development of handling systems can be complex and dicult. This chapter explains how to bring order and structure to the process. TheMFAsteps,showninFig.4are: 1. Classify materials 2. De®ne ¯ow unit 3. Determine data source 4. Extract data 5. Format and analyze 6. Calibrate ¯ows 7. Represent graphically. Material Handling and Storage Systems 609 Figure 2 Product volume/mode curve. Copyright © 2000 Marcel Dekker, Inc. These seven steps provide an understanding of the material ¯ows in the facility. The calibrated ¯ows are used to develop anity ratings. These initial steps are also the basis for subsequent evaluation of layout options and material handling system design. Step 1. Classify Materials. Most manufacturing and warehouse operations have a large variety of products and materials. Situations with 20,000 or more distinct items are not unusual. To analyze ¯ow or design a material handling system around so many individual items is not practical. Classi®cation reduces materials to a manageable number of items so that the classes then become the basis for determining ¯ow rates, con- tainers, and handling equipment. The initial classi®cations stratify materials for com- mon handling methods and container design. Weight, size, shape, ``stackability,'' and special features are 610 Wrennall and Tuttle Figure 3 Material varieties. Copyright © 2000 Marcel Dekker, Inc. de®ningcriteria.Figure5showsaclassi®cationbased on handling characteristics for a four-drawer cabinet. In addition, similarities in product, process sequence, and raw material are bases for grouping items that move over the same routes. Step 2. Identify Flow Units. Material ¯ow is measured in units of material over a unit of time and the analyst chooses appropriate units for both parameters. The time unit is usually a matter of convenience and depends largely on data availability. Typical examples are cases per hour, tons per day, pallets per week. Selection of the material ¯ow unit is more proble- matic. Where only one type of material moves, the selection is straightforward, for example, the bushel for a grain mill. But few facilities have only a single material or material type. A wide variety of size, shape, weight, and other handling characteristics must be con- sidered,asillustratedearlierinFig.3.Forexample, integrated circuits are tiny, delicate, expensive, and highly sensitive to electrostatic discharge (ESD), but the operations that use integrated circuits also use large metal cabinets. Between these extremes is a wide range of diverse items to move. Various items of the same size may have dierent handling requirements and costs. A resistor and an integrated circuit (IC) are very close in size. But resis- tors are moved in bulk, in ordinary containers, and without special precautions. The individual IC is sen- sitive to ESD. It requires an enclosed, conductive and expensive container. It may have a special tube or bag to further protect it. Humans may touch it only if they wear a grounded wrist strap and a conductive smock. Individual items or materials are seldom handled separately. Most items are in boxes, tote boxes, car- tons, bundles, bales or other containers. These contain- ers then are what need to be handled. But layout design requires a standard unit of ¯ow. This is the equivalent ¯ow unit (EFU) which should have the following char- acteristics: Applicable to all materials and routes Easily visualized by the users Independent of the handling method. The equivalent ¯ow unit should account for weight, bulk, shape, fragility, value, special conditions and other factors: Weight is a common unit for most materials and is usually available in a central database. Bulk, or density, relates weight and size. Overall dimensions determine bulk density. Shape impacts handling diculty. Compact regular shapes such as boxes stack and handle most easily. Round and irregular shapes stack with Material Handling and Storage Systems 611 Figure 4 Material ¯ow analysis. Copyright © 2000 Marcel Dekker, Inc. diculty. Long thin shapes (high aspect ratio) are awkward to handle. Fragility refers to the sensitivity of objects to damage. Fragility in¯uences handling diculty and expense; 100 lbs of sand and 100 lbs of glass- ware are very dierent handling tasks. Value for a wide range of objects and materials has little in¯uence. But high value or special security items such as gemstones require protection from theft, damage or loss. Special conditions that aect material handling dif- ®culty and expense are stickiness, temperature, slipperiness, hazard, and ESD sensitivity. As material moves through the production system, its aspects change and the handling eort, as measured by equivalent ¯ow units, can change drastically. For example: Bulk cellulose acetate ¯ake may be received and plastic ®lm rolls or sheets may be shipped. 612 Wrennall and Tuttle Figure 5 Material classi®cation summary. Copyright © 2000 Marcel Dekker, Inc. Tree trunks may be received and newsprint shipped. Bulk liquids and gases may be received but pharma- ceutical intravenous packs or bottles of tablets are shipped. Bauxite ore is received and aluminum ingots are shipped. Plywood is received, entertainment centers are shipped. Wood pulp and naphtha are received, chemicals, textiles, and plastics are shipped. What seems a minor change in the item sometimes brings a dramatic change in the equivalent ¯ow units. Figure 6 is a schematic ¯ow diagram that illustrates changes in ¯ow intensity as the material is processed for a four-drawer cabinet. Figure7isariverdiagramillustratingmaterial¯ow for all products in an entire plant. The diagram shows how ¯ow intensity increases after the material is painted and decreases after the parts are assembled. Painted sheet metal parts are easily damaged and di- cult to handle. Once assembled and packaged, the units become protected, compact, and stackable and their ¯ow in equivalent ¯ow units decreases dramatically for the same quantity and weight. When a decision is made on an equivalent ¯ow unit, convenience and familiarity often take precedence over accuracy. The primary purpose of this analysis is to rate ¯ow intensities into one of four categories. We use the vowel letter rating system A, E, I, and O. Accuracy of the order of Æ207 is therefore sucient. For this level of accuracy, the following procedure is used: Review potential data sources. Interview production and support personnel. Material Handling and Storage Systems 613 Figure 6 Equivalent unit ¯ow analysis. Copyright © 2000 Marcel Dekker, Inc. Observe operations. De®ne the equivalent ¯ow unit. Some examples of equivalent ¯ow units are pallets, bales, paper rolls, tote-boxes, letters, tons of steel, and computer cabinets. The analysis now requires factors to convert all materials into the equivalent ¯ow unit. Conversion may come from experience, work measurement, or benchmarking. An example from a jet engine overhaul facilityisgiveninFig.8. The graph converts item size to equivalent ¯ow units. Pallets and pallet-size containers were the most commonly moved items and the basis for most records. The equivalent pallet was, therefore, the most sensible equivalent ¯ow unit. The pallet container labeled ``PT'' is 1.0 equivalent ¯ow unit on the vertical scale. Its volume is 60 ft 3 on the horizontal scale. In this system, volume is the basis for equivalent ¯ow unit conversion. Several tote pans of dierent sizes are labeled ``2T,'' ``4T,'' ``8T,'' and ``8S.'' An assembled jet engine has a volume of about 290 ft 3 and an equivalent ¯ow unit value of 1.6 equivalent pallets. The relationship between volume and equiva- lent ¯ow unit is logarithmic rather than linear, which is not unusual. Jet engines are 4.8 times the bulk of a pallet load. On dollies they require only a small tow tractor or forklift to move. The cost and eort is about 1.6 times that for moving a pallet load. Additional factors can aect the logarithmic volume relationship. This accounts for dierences in density, shape or other handling modi®ers. Work standards can be used as conversion factors. The time and cost of moving representative items are calculated and compared, and become benchmarks for all other items in the facility, or the data might be the basis for a graphical relationship similar to the one illustrated previously in Fig. 8. Step 3. Determine Data Source. Almost every facility is unique with respect to the material data source. Products, volumes, and mix vary; practices are diverse, as are recording methods. Accuracy may be good, sus- pect, or demonstrably poor, and individuals who con- trol data sources may be co-operative or protective. This diversity necessitates extensive interviews with people who collect and compile the data. A good selec- tion of data source often makes the dierence between a dicult or an easy analysis. Here are some possible data sources: Process charts Routing sheets Material requirements database Routing database Direct observation Handling records Work sampling Schedule estimates Informed opinions. When selecting the data source, the analyst must also decide on the range of included items. All items should be used if their number is small or when computerized records make it feasible to do so. When a few products 614 Wrennall and Tuttle Figure 7 River diagram. Copyright © 2000 Marcel Dekker, Inc. represent the largest volumes and are representative of others, data from the top 20±30% should be used. Where groups of products have similar processes and ¯ows, a representative item might portray an entire group. When the product mix is very large and diverse, random sampling may be appropriate. Figure 9 illustrates data selection guidelines Process charts map the sequence of processes graph- ically; routing sheets often have much the same infor- mation in text form. With either source, each operation must be examined to determine in which SPU that operation will occur. This determines the route. From the product volume analysis or other information, the raw ¯ow is determined which is then converted to equivalent¯owunits,asillustratedinFig.10. This procedure is used directly if there are only a few products and where processes and ¯ows are similar and a single item represents a larger product group. For large numbers of items, process charts with a random sample are used. Material Handling and Storage Systems 615 Figure 8 Equivalent ¯ow units. Figure 9 Data selection guidelines. Copyright © 2000 Marcel Dekker, Inc. Most or all of the necessary information may exist in the databases of manufacturing requirements plan- ning and other production and scheduling information systems. It may be necessary to change the data to a format suitable for ¯ow analysis. Material handling records or direct observation are good sources for data. If material is moved by fork truck, for example, and a central fork truck pool keeps good records of pickups and deliveries, these records contain the necessary information. In direct observation, the observer follows products through various moves and operations. In this way both pro- cess and material ¯ow information are gathered simul- taneously.Thefrom±tochartofFig.11documents ¯ows obtained by direct observation. Several sources may be necessary to capture all ¯ows. For example, an MRP database may contain ¯ows for production items but not for scrap, mainte- nance, trash, or empty containers. These ancillary items are often signi®cant and sometimes dominant, particularly in high-tech industries. Step 4. Acquire the Data. After a data source is determined the data must be acquired. Computer- based data are accessed by information services. Other data sources may require considerable clerical effort. Direct observations or work-sampling derived data may require weeks to collect and process. Step 5. Format and Analyze the Data. Manual meth- ods can suf®ce for the entire material ¯ow analysis. However, computer-aided analysis is necessary for facilities with a wide product mix, process focus and a complex process sequence. Spreadsheet programs are suitable for most analyses. Database programs are sometimes better than spreadsheets because of their reporting and subtotaling capabilities. With computerized analysis, data can be entered as the project progresses. Initial data may come from downloaded information or manual collection and consist of product information such as names and part numbers and perhaps annual volume, weights, and routing. The analyst should consider ancillary uses for the database as well. The database may assist later in developing handling systems or determining storage areas. It might also be part of a group technol- ogy (GT) study for cell design. Figure12isanexampleofamaterial¯owreport used for the layout of a mail processing facility. Data came from a series of schematic material ¯ow charts, in turn derived from process charts, SPU de®nitions and a productvolumeanalysis,asshownearlierinFig.2. Fields 1 and 2 of Fig. 12 are SPU numbers which de®ne the ¯ow path for that entry. Fields 3 and 4 are descrip- tors corresponding to the SPU numbers. Field 5 is a type code; ®eld 6 is the equivalent ¯ow unit; ®eld 7 is the daily volume in pieces per day. All mail with the same 616 Wrennall and Tuttle Figure 10 Equivalent ¯ow units development process. Copyright © 2000 Marcel Dekker, Inc. [...]... warehouses often require high-density storage methods, such as drive-through racking 1.5.2 Storage Equipment The types of storage equipment available are almost as diverse as the types of containers and handling equipment The selection of both storage equipment and containers is interrelated 1.5.3 Analysis and Design of Storage Systems The design of storage systems should co-ordinate with the layout design of. .. York: Marcel-Dekker, 1990 G Salvendy Handbook of Industrial Engineering 2nd ed New York: Wiley-Interscience, 1992 ER Sims Planning and Managing Industrial Logistics Systems Amsterdam: Elsevier, 1991 JA Tompkins, JD Smith The Warehouse Management Handbook New York: McGraw-Hill, 1988 W Wrennall Requirements of a Warehouse Operating System In: JA Tompkins, JD Smith, eds The Warehouse Management Handbook. .. Handling April, 1986 Copyright © 2000 Marcel Dekker, Inc WK Hodson, ed Maynards's Industrial Engineering Handbook 4th ed New York: McGraw-Hill, 1992 M Hulett Unit Load Handling London: Gower Press, 1 970 AL Kihan Plant Services and Operations Handbook New York: McGraw-Hill, 1995 Modern Dock Design Milwaukee: Kelly Company, 19 97 W Muller Integrated Materials Handling in Manufacturing È IFS (Publications),... Itasca, IL: Putman Publishing Company, 1998, pp 74 79 2 W Wrennall, Q Lee, eds Handbook of Commercial Facilities Management New York: McGraw-Hill, 1994 3 HA Bolz, GE Hagemann, eds Materials Handling Handbook New York: The Ronald Press, 1958, pp 1.5±1.16 FURTHER READING CR Asfahl Robots And Manufacturing Automation, New York: John Wiley, 1985 A Carre Simulation of Manufacturing Systems, Chichester: John... can operate on a ®rst-in-®rst-out (FIFO) basis Floor stacking also requires strong, stable, stackable, and unbroken loads, illustrated in Fig 40 Pallet racks should be used when loads are unstackable or storage volume is too small for deep ¯oor stacking Double-deep racks achieve higher density storage but require a reach truck causing problems of access to the rear pallet Flow-through racks are used... controlling of material movement Included are manipulators, robots, positioning platforms, and transfers Also included are scales and weighing equipment, ¯oat controls, bin indicators, counters, and other control devices 400 Industrial Vehicles: this class includes all types of vehicles commonly used in and around industrial and commercial facilities Excluded are ``Motor Vehicles'' intended for over-the-road... now recognize that inventory often camou¯ages some form of waste The causes of waste are in the structure of the inventory systems The ultimate goal is to restructure and eliminate all storage of products Restructuring for minimum inventory is usually more fruitful than pursuing better storage methods, although compromises must be made and a requirement for some storage often exists 1.5.1 Stores Activities... and characteristics of material ¯ow Indeed, the ¯ow pattern dictates the shape or arrangement within a facility Figure 22 shows the basic ¯ow patterns: straight-through ¯ow, L-shape, U-shape or circular, and hybrids Figure 18 Locational ¯ow diagram (multiple lines) Copyright © 2000 Marcel Dekker, Inc With straight-through or linear ¯ow, material enters and exits at opposite ends of the site or building... conveyors use an I-beam or other shape as a monorail Carriers roll along the monorail with loads suspended underneath A chain connects the carriers and pulls them along In a powerand-free system, the chain and carriers are independent A disconnection mechanism stops the carrier Power-and-free systems oer more ¯exibility than standard monorails but at a much higher cost Recent designs of power and free... instead of forcing it.'' The two distinct categories of vibratory machines that are most often used in material handling systems are those for accomplishing induced vertical ¯ow and induced conveying 1.4 .7 Automatic Guided Vehicle Systems Automatic guided vehicle systems (AGVS) use driverless vehicles to transport materials within an operation 634 Wrennall and Tuttle AGV size can vary from small, light-duty . transported along pipes or ducts with the aid of a suspension agent, such as a gas or liquid. In just-in-time (JIT)/lean manufacturing the aim of batch sizes of one is to emulate continuous liquid or. arrangement of SPUs sets a large-scale ¯ow pattern (or nonpattern in some cases). Since layout design and material handling are inter- dependent, so is a discussion on the design of either or both of. sensitivity of objects to damage. Fragility in¯uences handling diculty and expense; 100 lbs of sand and 100 lbs of glass- ware are very dierent handling tasks. Value for a wide range of objects