1、PART II SolutionsModeling the Supply Chain.Modeling the Supply Chain(建構(gòu)供應(yīng)鏈方式)The examples in Chapter 4 show that even the simplest of systems can generatesurprisingly complex behavior. How, then, are managers to understand suchcomplex business systems as supply chains and manage them effectively? Th

2、eanswer, in a word, is modeling. The only way to understand complex systems isto construct simplified models of them, play with the models to see how theywork, and then apply what you learn to the real-world system. You may nothave thought about it this way, but you already do this with conceptual m

3、odelsall the time, even if the models are only in your mind. This chapter shows youhow to use these mental models more effectively, and it introduces two morepowerful kinds of modelsmathematical and simulation modelsthat you canuse as precision tools for prediction and control. You dont have to know

4、 howto build these more advanced models, but knowing what they are and when touse them is vital to managing a supply chain effectively.第4章的例子顯示了部使最簡單的系統(tǒng)，也會產(chǎn)生令人預(yù)料不到的複雜行為。那麼，經(jīng)理人要如何瞭解供應(yīng)鏈這麼複雜的商業(yè)系統(tǒng)，並且有效地管理它們呢？簡言之，答案就是建模(modeling)。瞭解複雜系統(tǒng)的獨一方法是去建構(gòu)它們的簡單方式，操控這些方式看看它們運作的法則，然後將所學(xué)到的應(yīng)用在真實世界的系統(tǒng)。他或許從來沒有這麼想過，但事實上他卻

5、不斷在運用觀念方式(conceptual models)思索真實世界的系統(tǒng)，即使這些方式只需在他的腦袋裡。本章將教他如何更有效地運用這些心智方式(mental models)，並且介紹兩種功能更強的方式數(shù)學(xué)方式與模擬方式，可作為預(yù)測與控制的精確工具。他不一定需求知道如何建構(gòu)這些進階的方式，但知道有這些進階方式存在，以及何時運用它們，對於有效管理供應(yīng)鏈?zhǔn)欠浅Ｖ匾摹?The Case for Models(方式之用處)Models are simplified representationsA model is nothing more than a simplified representat

6、ion of a real-world system. This representation can take a. wide variety offorms, including verbal explanations, whiteboard diagrams, mathematical equations, physical structures, and computer programs. All these models serve a common purpose: They take a system that may be hard to understand or dang

7、erous to manipulate, and they render it in a form that is easier to understand and safer to play with. 方式是簡化的表達方式方式(model)不過是一個真實系統(tǒng)的簡化表達方式。方式的表達有很多種方式，包含言語解釋、白板圖表、數(shù)學(xué)方程式、物理結(jié)構(gòu)以及電腦程式。一切類型的方式均具有一個共同目的：方式建構(gòu)會先選擇一個不易了解或其操作危險性的系統(tǒng)，再將這個系統(tǒng)以一種更容易了解，以及操作上更平安的方式呈現(xiàn)。.The Case for Models(方式之用處)Modeling aids the thre

8、e core processesModels help with all three of the key business processes described in Chapter 4: understanding, prediction, and control (Figure 5.1). Building a model of a system requires that you analyze the system to identify its key components, figure out how those components work, and then reass

9、emble them in a way that replicates the essential behavior of the system. That basic sequence of analysis and synthesis is the surest way to understand complex systems. Its also the method underlying most scientific discoveries, engineering solutions, and business innovations.建模協(xié)助三項中心程序方式有助於第4章描畫的三個

10、關(guān)鍵商業(yè)程序(three of the key Business processes)：瞭解、預(yù)測及控制(見圖5.1)。建立一個系統(tǒng)的方式需求他去分析此系統(tǒng)的關(guān)鍵元件，釐清這些元件如何運作，並用一些方法重組這些元件，以複製此系統(tǒng)的必要行為(essential behavior)。分析與綜合的基本順序(瞭解(understanding)、預(yù)測(prediction)及控制(control)，是瞭解複雜系統(tǒng)的最好方法。此方法也是許多科學(xué)發(fā)現(xiàn)、工程解答，以及商業(yè)創(chuàng)新當(dāng)中所採用的方法。.The Case for Models(方式之用處)Modeling aids the three core pro

11、cessesModeling SupplyChains.The Case for Models(方式之用處)Models generate valuable predictionsOnce you have assembled a model, you can use it as a test bed to generate predictions about how the system it represents would behave under a variety of conditions. Would building a new warehouse in Omaha reduc

12、e transportation costs as much as expected? Could the current supply chain support a 15% increase in demand? What would be the impact on cash flow of extending better credit terms to key customers? Its possible to answer what if“ questions such as these by changing the real-world system, but models

13、provide answers faster, less expensively, and with a lot less risk to the company. Predictions generated by business models, in turn, increase your understanding of the real-world system, and you can use that increased understanding to further improve the quality of the model and its predictions.方式可

14、產(chǎn)生有價值的預(yù)測一旦他組合出一個方式，他可以運用此方式作為一個測試工具，產(chǎn)生關(guān)於方式所代表的系統(tǒng)在不同情況下的預(yù)測。在奧馬哈(Omaha)建立一個新倉庫所減少的運輸本錢與預(yù)期能否一樣？目前的供應(yīng)鏈可以支持15的需求增加量嗎？提供較好的信譽條件給關(guān)鍵顧客，將對現(xiàn)金流量帶來什麼影響呢？此類假設(shè)則問題(what if guestjons)可透過改變真實世界系統(tǒng)來回答，但是方式卻可更快速地提供答案、較不昂貴，且對公司而言風(fēng)險較小。由商業(yè)方式產(chǎn)生的預(yù)測將可添加我們對真實世界系統(tǒng)的瞭解，因此他可以運用這些添加的瞭解來改善方式及其預(yù)測的品質(zhì)。.The Case for Models(方式之用處)Good m

15、odels make for good decisionsTo see how important predictions are to helping you improve your supply chain, think back to the example in Chapter 3, in which a customer and a supplier wanted to create a shared win by reducing their total inspection costs. The tradeoff curve shown in Figure 3.11reveal

16、ed that the best arrangement was for the supplier to spend a dollar more on inspection, allowing the customer to spend four dollars less. Where did this tradeoff curve come from? It could only be the result of a model that took into account the operations required for quality assurance, the cost of

17、those operations, and their net effects on quality. In practice, a company wouldnt actually draw this tradeoff diagram once it had the model; it would simply use the model to find the lowest-cost solution. The only reason for drawing the diagram would be to help people understand why the new inspect

18、ion program is a win for both companies.良好方式產(chǎn)生良好決策要知道預(yù)測對供應(yīng)鏈改善有多重要，回想一下第3章的例子，一個顧客與一個供應(yīng)商想藉由減少總檢驗本錢來創(chuàng)造雙贏(shared win)。圖3.11互抵效應(yīng)曲線顯示，最好的安排是供應(yīng)商多花費1元的檢驗本錢，顧客則可少花費4元。此互抵線來自於何處？它只能夠來自於一個方式的分析結(jié)果，該方式考慮到品質(zhì)保證所需的作業(yè)、這些作業(yè)的本錢，以及它們在品質(zhì)上的淨(jìng)效果。實際上，一家企業(yè)不會真正畫出此互抵線圖，一旦他建構(gòu)了模式；他只需運用此方式找出最低本錢解答即可。繪製此圖形的獨一理由乃在幫助大家瞭解，為何新的檢驗方案是個

19、雙贏的結(jié)果。.The Case for Models(方式之用處)Models are also used for controlModels are also used to control real-world systems, as shown on The right of Figure 5.1. This use of models is less obvious than the other two because the models used in control are usually implicitthat is, they are embedded in the des

20、ign of business systems, but are never communicated to the people who own and operate those systems. This problem is particularly acute in the case of software. All the various kinds of supply chain software described in the next chapter are based on very specific models about how production, distri

21、bution, replenishment, sales, and other business processes are carried out. Unfortunately, most of the companies that buy this software have no idea what those models are until after they install the system and discover that the embedded models dont support the way they do business. The impact of th

22、is mismatch can range from a minor nuisance to a complete failure of the system, which is certainly a devastating way to learn about the importance of business models.方式也運用於控制方式也運用於控制(control)真實世界系統(tǒng)，如圖5.1右邊所示。方式在控制上的運用比其他兩個功能較不明顯，那是因為運用於控制上的方式通常是內(nèi)隱的(implicit)；亦即，它們被置入於商業(yè)系統(tǒng)的設(shè)計中，但是從來沒告知那些擁有與操作這些系統(tǒng)的人。這

23、個問題在軟體案例中特別地嚴(yán)重。在下一章介紹的各種不同類型供應(yīng)鏈軟體均植基於非常特定的方式，有關(guān)於製造、配銷、補貨、銷售和其他商業(yè)程序如何被執(zhí)行。不幸地，絕大部分企業(yè)在購買此軟體時，根本不知道有哪些方式，直到他們安裝系統(tǒng)後，才發(fā)現(xiàn)軟體所置入的方式無法援助他們的商業(yè)運作。這項不搭配的衝擊能夠從小至一個芝麻蒜皮的問題，甚至大到系統(tǒng)的完全失敗都有能夠，這當(dāng)然是一種可怕的方式來瞭解商業(yè)方式的重要性。.The Case for Models(方式之用處)There are three basic types of ModelsBusiness models come in a wide variety o

24、f forms, but most of them fall into one of thethree broad categories shown in Figure 5.2. Conceptual models use diagrams anddescriptions to represent a business system. They can be created on white-boards,computer screens, or the backs of envelopes, and they provide simple, familiarstructures for re

25、asoning about the business. Mathematical models represent a business in terms of formulas and procedures, and they are solved by evaluating those formulas or procedures under a particular set of assumptions. Simulation models use software objects to represent the components of a business, and they a

26、re solved by running the model to see what happens when the objects interact with each other. Mathematical and simulation models are often referred to as formal models because they have strict forms and generate numerical predictions, in contrast to the informality of conceptual models.方式的三種根本類型商業(yè)方式

27、有多種方式，但大多數(shù)屬於圖5.2所示三種主要類型之二種。觀念模式(conceptual models)運用圖形與描畫來表達一個商業(yè)系統(tǒng)。觀念方式可以在白板、電腦螢?zāi)换蚴切欧夥疵嫔蟿?chuàng)造，而且它們提供了解商業(yè)系統(tǒng)之簡單、熟習(xí)的結(jié)構(gòu)。數(shù)學(xué)方式(mathematical models)以公式(formulas)與運算程序(Procedures)代表一個商業(yè)系統(tǒng)，這些數(shù)學(xué)方式在一組特殊假設(shè)條件下，藉由估算公式與運算程序求得解答。模擬方式(simulation models)運用軟體物件(software objects)來表達一個企業(yè)的元件components)，模擬方式藉由跑(running)方式來瞭

28、解不同物件互動後，所發(fā)生的情形來尋找解答。數(shù)學(xué)與模擬方式通常被稱為正規(guī)方式(formal models)，因為它們有很嚴(yán)格的方式且會產(chǎn)生預(yù)測數(shù)值，與非正規(guī)化的觀念方式構(gòu)成對比。.The Case for Models(方式之用處)Three kinds of Models.The Case for Models(方式之用處)Each type offers unique AdvantagesThe distinctions among these three kinds of models are not hard and fasthybrid forms are common. But th

29、e three types do represent three fundamentally different approaches to modeling, and each offers a unique set of capabilities and limitations. Because conceptual models are the easiest to build and understand, they are the best choice for achieving a shared understanding of the supply chain, particu

30、larly when managers are involved in the modeling process. Mathematical models are the most powerful, and they are best used to predict and optimize the performance of the chain. Simulation models are the most flexible, and they should be used to study the behavior of a model under the most realistic

31、 business conditions.每種方式均有其獨特優(yōu)勢這三種方式之區(qū)別並不困難而且非常快 混合型態(tài)(hybrid forms)很常見。不過這三種類型確實代表三種根本上不同的建模方法，而且每一類方式均有其獨特的才干與限制。因為觀念方式最容易建構(gòu)與瞭解，對於達成供應(yīng)鏈的共同性瞭解，特別是在經(jīng)理人參與建模過程時，它們是最正確選擇。數(shù)學(xué)方式是最具才干的，它們運用在供應(yīng)鏈績效的預(yù)測和最正確化上是最好的。模擬方式是最有彈性的，它們應(yīng)該運用於瞭解在最真實企業(yè)環(huán)境下的商業(yè)方式行為。.The Case for Models(方式之用處)Choosing the type is a management

32、 decisionAs a manager, you dont need to know how to use mathematical models and simulations; these formal models are usually implemented in software, and specialized skills are required to set them up and run them. But you do need to know how to use conceptual models because you are already applying

33、 them, whether well or badly, and you need to know what to expect of the other two types and when to trust their output. The next three sections provide a quick tour of the three kinds of models, and the final section offers some guidelines for using them to solve supply chain problems.選擇方式類型是個管理決策作

34、為一個經(jīng)理人，他不需求知道如何運用數(shù)學(xué)方式與模擬方式；這些正規(guī)方式通常包含在軟體中，方式之設(shè)定與運用需求具備特殊技藝。但他需求知道如何運用觀念方式，因為他已經(jīng)在應(yīng)用它們了，不論是好或壞；而且他需求知道對其他兩類方式應(yīng)有的等待，以及何時去信任它們的產(chǎn)出。以下三節(jié)提供三種方式類型的快速導(dǎo)覽；最後一節(jié)則提供運用此三類方式解決供應(yīng)鏈問題的一些指導(dǎo)方針。.Conceptual Models(觀念方式)Conceptual models are the simplestThe conceptual model is by far the simplest of the three types. This

35、sort of model is basically a description of a business system and is usually expressed as some combination of diagrams and explanations. To a large extent, the format depends on the experience of the modelersthose with the most training in modeling usually rely on detailed diagrams with formal notat

36、ion to reduce ambiguity. By contrast, those with little or no training tend to express their models as verbal descriptions mixed with stories about how the business worksstories that can often be formalized as scenarios. Although generally less precise than diagrams, descriptions and scenarios often

37、 capture the nature of the business in a way that formal diagrams cannot. The best conceptual models are usually a mix of diagrams, descriptions, and scenarios.觀念方式是最簡單的方式觀念方式是三種方式中最簡單的。這類方式根本上是對一個商業(yè)系統(tǒng)的描述，通常是以一些圖形與說明的組合性表達。普通來說，其方式取決於建模者的經(jīng)驗具備嚴(yán)格訓(xùn)練的建模者通常採用正式符號的詳細圖形，以減少模糊性。相較之下，較少或是沒有訓(xùn)練的建模者則偏向於以口頭敘述，摻雜

38、一些關(guān)於商業(yè)運作的故事來表達他們的方式這些故事可以進一步轉(zhuǎn)變?yōu)橐恍┣榫?scenarios)。雖然通常比圖表不詳實，描畫和情境時常以較簡易方式表達商業(yè)的本質(zhì)，是正規(guī)方式所無法做到的。最好的觀念方式通常是圖形(diagrams)、敘述(descriptions)和情境(scenarios)三者的相結(jié)合。.Conceptual Models(觀念方式)The goal is o facilitate communicationRegardless of how you express a conceptual model, the key is to find the rightbalance b

39、etween precision and ease of communication. For systems analyststrained in the use of entity relationship (ER) diagramming, formal ER diagramsand detailed scenarios May be just the right tools. For managers who have never engaged in business modeling before, the right balance may be a combination of

40、 simple diagrams and informal explanations. But even with managers, some conventions are necessary to make the diagrams and explanations make sense. Otherwise the output of the process may contain more myth than model.觀念方式的目標(biāo)是在促進溝通不論他如何表達一個觀念方式，關(guān)鍵在於找出介於精確和容易溝通之間的平衡點。對於受過實體關(guān)係繪圖方法(entity relationship

41、(ER) diagramming)訓(xùn)練的系統(tǒng)分析師而言，正式ER圖形以及詳細情境描畫也許是正確的工具。對於不曾投入商業(yè)建模的經(jīng)理人而言，好的平衡點也許是一種簡單圖形和非正式說明的組合。但即使是經(jīng)理人，仍需遵照一些規(guī)範(fàn)才干使圖形與說明讓人了解。否則，程序產(chǎn)出能夠比方式本身更難讓人了解。.Conceptual Models(觀念方式)This book uses convergent engineeringThe diagrams in this book generally follow the conventions of convergent engineering,a modeling t

42、echnique I developed specifically to help managers formulate useful business models. In this approach, a business system consists of three basic kinds of objects: organizations, processes, and resources. As shown in Figure 5.3, each of these objects plays a different role in the model, and the three

43、 relate to each other in ways that both constrain the model and make it more understandable at the businesslevel. Briefly put, organizations own resources and execute processes; processes consume one set of resources and generate another set; and resources are the source of all cost and value in the

44、 system. There is much more to the approach than this, of course (see my earlier book, Business Modeling with Object Technology), but this one-sentence summary illustrates what I view as an appropriate level of formalism for managers, and it should make the illustrations more meaningful to you as we

45、ll.本書採用收斂工程本書圖形主要採用收斂工程(convergent engineering)的慣例，乃筆者發(fā)展出的一種建模技術(shù)，特別有助於經(jīng)理人建構(gòu)有用的商業(yè)方式。在此方法中，一個商業(yè)系統(tǒng)是由三種根本物件(objects)所構(gòu)成：組織(organizations)、程序(processes)和資源(resources)。如圖5.3所示，這些物件在此方式中個別扮演不一樣的角色，三者之相互關(guān)係在方式中彼此牽制，並讓此方式在商業(yè)水準(zhǔn)上更加容易了解。簡單來說，組織擁有資源並且執(zhí)行程序；程序耗費一組資源並產(chǎn)生另一組；而資源是一切本錢與價值在此系統(tǒng)中的來源。當(dāng)然，收斂工程的內(nèi)容遠較此G的介紹為多(請參考

46、筆者早期的著作：(Business Modeling with Object Technology)，但上述總結(jié)闡明筆者所認為經(jīng)理人應(yīng)具備的正規(guī)建模水準(zhǔn)，並且這種建模方法的表達方式也應(yīng)該會添加他對方式的了解程度。.Conceptual Models(觀念方式)This book uses convergent engineeringOrganizations,Processes, andResources.Conceptual Models(觀念方式)Group modeling is the most effectiveConceptual models can be developed b

47、y individuals, but for systems that cross organizational boundaries, as supply chains inevitably do, the best approach is to assemble a team of representatives from all the groups involved and hammer out the model together. Many software tools have been designed to support this group design process,

48、 but low-tech tools are often the most effective. Personally, Ive always gotten the best results from a combination of whiteboard diagrams and 5X7 index cards. Each card represents one of the organizations, processes, or resources required for the model, and participants take turns role-playing thes

49、e objects as they interact in the operation of the business. The resulting process is highly engaging, often contentious, and alwayseducational as participants discover that each has a radically different understanding of how the business actually works. Once the group has assembled a consensus mode

50、l out of its various conflicting perspectives, it has a solid foundation on which to build a better system.集體建模是最有效的觀念方式可以由個人來建構(gòu)，但對於橫跨組織界線的系統(tǒng)來說，如供應(yīng)鏈之例，最好的方法是從一切相關(guān)單位選出代表，組成一個任務(wù)團隊，一同完成方式之設(shè)計。許多軟體工具已經(jīng)被設(shè)計用於援助此類集體設(shè)計(group design)程序，但低技術(shù)性工具(low-tech tools)經(jīng)常是最有效的。就筆者而言，筆者總是從綜合運用白板繪圖與5X7索引卡當(dāng)中獲得最好的結(jié)果。每一張卡片代表

51、方式中的一個組織、程序或資源，由參與者輪流扮演這些物件的角色，並進行相對應(yīng)商業(yè)作業(yè)上的互動。最後產(chǎn)生的程序是人員高度投入且時常爭論不已所得，非常其有教育意義，因為參與者在過程中發(fā)現(xiàn)，每個人對企業(yè)實際運作方式上竟然有完全不同的理解。一旦這個團隊透過這樣的衝突過程完成具有共識的方式後，一個堅實的基礎(chǔ)即已建立，並可在此基礎(chǔ)上發(fā)展更好的系統(tǒng)。.Conceptual Models(觀念方式)Conceptual models are poor at predictionAlthough conceptual models form the basis for understanding systems,

52、 they areof little value in prediction and control. It should be clear from the precedingchapter that even the simplest models can produce surprising interactions assoon as two or more components are hooked together, and our minds are simplynot equipped to extrapolate the effects of these interactio

53、ns. When we do try topuzzle out the behavior of a system, most of us tacitly assume that all therelations involved are linear. For reasons that psychologists are still teasing out,it is extremely difficult for us to extrapolate the behavior of nonlinearinteractions, so we naturally tend to work with

54、in our limitations and oversimplify real-world relationships. Going beyond these limitations requires us to turn to more powerful kinds of models.觀念方式的預(yù)測才干非常薄弱雖然觀念方式是瞭解系統(tǒng)的基礎(chǔ)，它們在預(yù)測與控制上的價值卻很小。從前一章中我們應(yīng)該瞭解，即使是最簡單的方式，當(dāng)兩個或兩個以上的方式元件要連結(jié)在一同時，都能夠產(chǎn)生出令人驚訝的互動效果，而且我們的心智沒有足夠裝備去看出這些互動的效果。當(dāng)我們試著去找出一個系統(tǒng)的行為時，我們大多下意識地假定

55、一切的關(guān)係都是線性的?；兑恍┬乃紝W(xué)家仍舊會訕笑(teasing out)的緣由，人類在推斷非線性互動行為時存在非常高的困難度，因此，我們很自然地傾向於在這種限制條件下任務(wù)，並將真實世界的關(guān)係看得過於簡單。欲跳脫出這些限制條件，我們必須更改方向?qū)で蟾咝艿姆绞筋愋汀?.Mathematical Models(數(shù)學(xué)方式)You already use mathematical modelsRemember those word problems you hated as a kid? They went something likethis: If a boat moving upstream

56、 in a river flowing at 2 miles an hour takes 4hours to travel 3 miles, how many people were in the boat? These exerciseswere designed to teach you How to generate and apply mathematical models.And despite this early training, you do use mathematical models today. You just dont do it formally.他已經(jīng)在運用數(shù)

57、學(xué)方式了他還記得小時候他所討厭的代數(shù)問題嗎？例如，假設(shè)一條船每小時往上游移動2英哩，花了4小時游覽了3英哩，有多少人在這條船上？這些練習(xí)題是設(shè)計來教育他如何創(chuàng)造與應(yīng)用數(shù)學(xué)方式。雖然這是學(xué)生時代的訓(xùn)練，我們今日還是在運用數(shù)學(xué)方式，只是沒有很正式地應(yīng)用它們而已。.Mathematical Models(數(shù)學(xué)方式)Production costing use a linear modelFor example, suppose your boss asks you how much it would cost to run batches of 1,000, 2,000, or 3,000 audi

58、o CDs. You know that it costs $1,000 to set up the run and a dollar to make each CD once the run begins, so your total cost would be $ 1,000 plus $ 1 times the number of CDs, giving you $2,000, $3,000, and $4,000 for the three quantities. In working up those numbers, you used one of the most common

59、mathematical models in all of businessa linear model. In effect, your model predicts a linear relation between cost and the number of CDs produced, as shown in Figure 5.4.生產(chǎn)本錢分析運用的是一個線性方式舉例來說，假設(shè)他的老闆問他分1,000、2,000、3,000幾個批量生產(chǎn)音樂光碟，各批量所需的生產(chǎn)本錢為何？他知道生產(chǎn)設(shè)置的固定本錢為1,000美元，並且每生產(chǎn)1片光碟要花費1美圓，所以他的總本錢將會是1,000美圓加上1美

60、圓乘上光碟生產(chǎn)數(shù)量，得出三種批量的生產(chǎn)總本錢分別為2,000美元、3,000美圓、4,000美圓。在計算這些數(shù)值時，他已經(jīng)在運用商業(yè)上最常應(yīng)用的一種數(shù)學(xué)方式線性方式(linear model)。實際上，他的方式預(yù)測生產(chǎn)本錢和光碟產(chǎn)量為一種線性關(guān)係，如圖5.4所示。.Mathematical Models(數(shù)學(xué)方式)Production costing use a linear modelProduction costing by Volume.Mathematical Models(數(shù)學(xué)方式)Mathematical models are actually systemsAs this exa