The dimensions of process knowledge

The dimensions of process knowledge

RESEARCH ARTICLEKnowledge and Process ManagementTable 2 APQC classification of organizational processes (APQC, 2004)Operating processesDevelop vision and strategyDesign and develop productsMarket and sell productsDeliver products and servicesManage customer serviceManagement/support processesDevelop and manage human capitalManage information technology and knowledgeManage financial resourcesAcquire, construct and manage propertyManage environmental health and safetyManage external relationshipsManage improvement and changedesign, manufacturing/service, financial andaccounting, administrative, legal and managerial.Examples of these can be found in Tables 1 and 4.The classification covers all manner of aggregatedorganizational activity. Process P1 in Table 1 is anexample of a manufacturing process, P2 an exampleof a financial/accounting process and P3 an exampleof a design process. The reader will note alsothe rationale for classifying financial/accountingprocesses as operational processes rather thanadministrative or support processes. In a bank, lendingis an operational process by our classification,whereas this would be classified as an administrativeor support process in APQC.ORGANIZATIONAL LEARNINGAND PROCESS KNOWLEDGEDifferent frameworks have been advanced to characterizethe knowledge management cycle in organizations(Demarest, 1997; Pearlson, 2000; Ruggles,1998). The major activities in the cycle include identification,generation, codification and transfer. Thepreliminary stages of identification and generationare concerned with the acquisition of knowledge.Sources of knowledge could be internal or external,including experienced employees, consultants,experts and trade reports. The relevant knowledgeis identified and generated through familiar techniquessuch as discussion groups, presentations andshared workspaces. Since we are proposing a frameworkto aid in knowledge organization, we areconcerned with the first three stages of the cyclein this research.The generation of process knowledge differsfrom the generation of tacit knowledge held byemployees. The latter is generated through discussiongroups, online conferencing etc., while explicitprocess knowledge is generated as a result of consciousmanagement and monitoring of processes.Table 3 summarizes these differences. Organizationsdepend on process improvements for productivityincreases. Accordingly, they are continuouslyfine tuning the parameters of the process such aschanging the order of activities, temperature, pressureand ingredients, to increase throughput, qualityand efficiency. The process of monitoringand making changes can be characterized asObserve, Analyze, Design, Implement (Kim,1993). Thus, when changes are implemented to aprocess, results are observed, analyzed, andimprovements are made. This process of learningis said to be single-loop learning (performance loop),which occurs when an employee observes processoutputs and makes modifications to improvethem. Double-loop (relevance loop) learning occurswhen an employee questions the beliefs andTable 3Differences between personnel and process knowledgeKnowledge domain ! Personnel knowledge Process knowledgeType of knowledge Mostly tacit Mostly explicit aDegree of formality Informal: based on employee experience Formal: based on organizationalexperienceMethod of generation Interviewing, discussion groups, etc. Process management: Observe, Analyze,Design, implementExtent to which it can be codified Not to a great extent Explicit part, provided context ispreserved ba As stated earlier, one aspect of process knowledge, namely ‘performance knowledge’, shares the same characteristics as ‘personnelknowledge’ referred to in this table.b See Wensley and Verwijk-O’Sullivan (2000).68 C. S. Amaravadi and I. Lee

Knowledge and Process Managementassumptions behind the process set-up and makesfundamental changes to its structure (Davenportand Beers, 1995; Hackbarth and Grover, 1999). Forinstance, consider a computer firm that has traditionallyleased equipment to small and mid-sizedcompanies. For these types of clients, leasing is anexpedient method of reducing capital costs. Theleasing process could involve activities such asidentifying the client’s needs, preparing a contract,and delivering the equipment. An employee observingthe process could suggest improvements tothe manner in which the process is carried out. Ifthe employee observes that salespersons need towrite down the client needs, go to the office andhave a contract made out, he/she could suggestthat both activities be carried out simultaneously.On the other hand, observing the prices on newand used equipment, and the willingness of clientsto spend on information technology, he/she couldquestion the rationale behind the leasing processand could suggest fundamental changes to the processitself, namely, providing direct sales of equipment.Single-loop learning in this case involvesobserving and learning from process performance,and results in changes to process parameters such aschanging the order of activities. Double-loop learninginvolves learning about the process in relationto its environment and making a radical change tothe process, namely the inclusion of direct sales.Both types of learning activities require and generatea considerable amount of knowledge, the characterizationof which is one of the tasks at hand.RESEARCH ARTICLEDIMENSIONS OF PROCESS KNOWLEDGEKnowledge, according to Davenport and Prusak(1998), is ‘a fluid mix of framed experience, values,contextual information, and expert insight that providesa framework for evaluating and incorporatingnew experiences and information’.Knowledge is contextual and includes an actionablesummary and interpretation of experience.Similarly, process knowledge is also experiential,contextual and actionable. A process is the resultof institutionalization of practice as pointed outearlier and process knowledge is a valuable byproductof this process. It is contextual since it is difficultto characterize outside of the process. It isembedded in structure, training, management andtechnologies. Thus process knowledge is defined ascontextual, experiential, value laden and insightfulinformation about a process, including how it isconfigured, how it is coordinated, how it is executed,what outputs are desirable and whatimpacts it has on the organization. It is actionable,since the knowledge can be used for trainingemployees, communicating best practices or foreffecting improvements. The best way of characterizingthis knowledge is to use the Observe, Analyze,Design, Implement framework (Kim, 1993).We have re-ordered these as Design, Implement,Observe, Analyze, but rather than use these labels,we will use process-specific labels as illustrated inFigure 2. These labels are popular in the workflowliterature (for an example, see WFMC, 1999). TheFigure 2Dimensions of process knowledgeThe Dimensions of Process Knowledge 69

RESEARCH ARTICLEprocess learning cycle starts with defining the process,then proceeds through execution, observationof outcomes and feedback.The dimensions of knowledge are associatedwith the modified learning cycle. The definitionor specification of processes creates structuralknowledge. It is then executed for a particularinstance (example, John Doe’s loan application).Process execution in our view requires a varietyof types of knowledge. Specifically, employeesneed to be trained and coordinated to execute theprocess. The knowledge required to train andcoordinate them becomes part of process knowledge.Some processes require employees to usetools. Software testing using a testing environmentis a case in point. The tool allows employees to rundifferent modules on different test cases and itallows them to record and analyze the results.When employees carry out processes, they accumulateknowledge and this becomes one of the tacitdimensions of process knowledge, namely performanceknowledge. Managerial processes typicallyinvolve multiple participants, framing of context,complex sets of trade-offs, and other careful discriminationswhich are also present in other complexprocesses (Garvin, 1997). The record of theseactions is what we are characterizing as thediscourse dimension.When a process has executed, it produces an outcome,and hence the results dimension is alsoincluded (please refer to Figure 2 again). In certaintypes of processes such as manufacturing processes,it is important for the results to meet qualityrequirements. Examples are tolerances, yields, anddefects. Non-manufacturing processes would alternativelyhave certain objectives such as being withinbudget or hiring within a time period, which havethe same significance as quality in manufacturingprocesses. Finally, the execution of a process couldhave impact on the same process or on other processes.A hiring plan developed in an organizationcould result in employees being hired. Feedbackcould affect the structure or performance or otherprocesses. Thus the dimensions of process knowledgeare: structural, personnel/coordination,performance/tools, discourse, results, quality/objectives, and impacts/implications. These areformally discussed next.StructuralThe structural dimension is concerned with configurationsof a process, particularly the orderings ofactivities which characterize organizational processes(Malone et al., 1999). It encompasses thesequence in which activities have to be performed,Knowledge and Process Managementthe inputs and outputs that they have, the constraintsunder which they are carried out, and themanner in which these can be changed to optimizethe process. There are various methodologies formodeling the process structure and the literatureon this is very extensive. Such methodologiesinclude Petri nets, variations on state transition networks,program specification techniques, transactionmodels, UML, logic, and frames (Amaravadi,2004). SAP, for example, has a methodology formodeling business processes which it refers to asthe event process chart. The methodology includesconstructs as well as icons to represent interactionsbetween events, functions, processes, and organizationalunits (Curran et al., 1997).The description of a process enables formation ofmental models and is often the main vehiclefor knowledge sharing (Leppanen, 2001). Processdescriptions are vital in manufacturing processessince the parameters and constraints are stringent.In the manufacture of chemical compounds,for instance, the reaction times, the temperatureand pressure, composition of input materials, andthe type and composition of catalyst used need tobe closely adhered to or the firm risks losingthe entire batch. On the other hand, in design andmanagerial processes, there is a greater cognitivecomponent and the opportunity for variations instructure within different groups. Organizationsattempt to experiment with process configurationsin order to increase throughput, reduce cycletime, or increase efficiencies. The record and rationalefor these changes are a valuable derivative ofthe learning process and essential to knowledgesharing.Personnel and coordinationIn processes where manual intervention is necessary,the company’s employees have a very importantbearing on the output. ‘The operators must bevery flexible in their problem solving behavior...very ingenious in using all sorts of availableknowledge when applicable and their diagnosticperformance increases with experience’ (Leppanen,2001, p. 579). Since processes are typically handledby multiple employees, the personnel and coordinationdimensions refer to the training and managementthat are necessary for the process to achieveits desired result. Thus it includes issues such as:What are the process expertise requirements?How many individuals are required? How oftenshould they meet? What criteria should be usedto evaluate them? How should they be trained?For how long should employees be trained? Howshould they be coordinated? What development70 C. S. Amaravadi and I. Lee

Knowledge and Process Managementprograms have been carried out and what are theresults of these programs?Performance and toolsKnowledge is both required and generated as aresult of process execution. The performancedimension is concerned with knowledge associatedwith the execution of the process and tools used.Included under this rubric is knowledge aboutthe factors which affect the efficiency and throughput,the type of problems that arise and their resolution,the support tools and their peculiarities.One example of this type of knowledge (in a fastfood restaurant) is a method of distributing pepperonion deep-dish pizzas so that it overcomesthe problem of ‘clumping’, where all the pepperonigathers in the middle. The solution to even distributionis to arrange the pepperoni in the form ofspokes, on the pizza, before placing them in theoven (Argote, 1999, p. 91). Overcoming problemsassociated with process execution is a significantcomponent of performance knowledge. Since toolsare utilized to execute processes, these are alsoassociated with performance knowledge. Tools areartifacts required or utilized in process execution.Tools can be physical tools such as jigs, fixtures,and testing apparatus, or conceptual tools such asengineering methodologies. Tool knowledge isconcerned with how to operate the tools, its settingsand idiosyncracies. At the outset of this section,we discussed the example of a softwareenvironment to carry out testing. To carry out thetesting, an operator has to know the settings, thedifferent test options and the order in whichthe inputs are supplied. Thus performance/toolknowledge includes knowledge of tools, alongwith knowledge of how to execute the processand troubleshoot problems.DiscourseCertain types of processes such as strategy formulationand design are iterative and involve a considerableamount of negotiation and discourse, toobtain ideas, surface problems, obtain additionalinformation, resolve issues, and arrive at a consensus.The discourse dimension, sometimes calleddue process, refers to the meandering process ofarriving at decisions (Hewitt, 1986, Gerson andStar, 1986). It is time consuming, and involves multipleparties and trade-offs/compromises. Hiringthe CEO of a company is such an example since itis a lengthy process involving the board, the personneldepartment and top management of thecompany, with each group having different views.RESEARCH ARTICLEThe discourse process generates knowledge aboutthe history leading up to a decision: the rationalebehind decisions, the time frames, the key actors,the alternatives considered, and the compromisesthat were made (Hewitt, 1986; Gerson and Star,1986). This type of knowledge is necessary to evaluateor trouble-shoot processes or to make deeperadjustments (incorporate double-loop learning) tothem.ResultsThe results dimension concerns two types of knowledge:the outcomes of a process being executed andresults concerning its effectiveness. For instance, ina loan situation, the amount of the loan issued asassociated with the type of customer is an exampleof the former type of knowledge. Actually, it wouldbecome knowledge only when analyzed over a periodof time over a large number of customers(Hackbarth and Grover, 1999). Patterns such asthe type of customers and average loan amountscan then be discerned. The second type of knowledgeconcerns results of process measures, i.e. ‘performanceloop’ type information. In the loansituation, these can include the number of ‘touchpoints’ (the number of times the bank handles aparticular application), the amount of time it tookto process the loan, the number of write-offs (dueto non-payment), etc.Quality and objectivesFor manufacturing and service processes, productquality is one of the important, albeit intangible,outputs of a process. The quality dimension is concernedwith quality of the process and its outcome.It encompasses knowledge about quality indicators,their current and target values, and techniquesto improve quality. Quality indicators include suchthings as timeliness, cost, and quantity. Thisdimension has some overlap with the tools andperformance dimension. For non-manufacturingprocesses, quality may not be relevant and thereforewe have also included the label of objectivesto encompass the requirements to be met byadministrative and managerial processes. It caninclude such things as minimizing claim amounts(in claims processing) or hiring a CEO by a certaindate (in a hiring process).Impacts and implicationsProcesses are typically interlinked with otherprocesses such that changes to one have importantimplications for other activities within theThe Dimensions of Process Knowledge 71

RESEARCH ARTICLEorganization. In fact, these are often the critical processesin the organization (Lientz and Rea, 1998).Design and managerial processes are cases in pointsince they are concerned with key decisions thatdrive other OP. The redesign of an engine couldrequire retooling of assembly plants, modificationto components that are purchased, and perhapschanges in supplier relationships as well. Similarly,when a company decides to issue stock, it is obligatedto inform the Securities and Exchange Commissionand its shareholders. The implicationsdimension is thus concerned with implications fororganizational action. As with the results dimension,there are two types of implications correspondingto the two types of learning. The resultsof a process could have implications for makingadjustments to the process or for making changesto other activities. It should be evident that thisdimension comes into play when modifying existingOP or designing new OP based on existingprocesses.AN ILLUSTRATION OF THE DIMENSIONSOF PROCESS KNOWLEDGEThe following mini-case describes the process ofdesigning components (such as suspensions, transmission,and engines) for a heavy equipment manufacturer(HEM) located in the Midwest. HEM 6manufactures a variety of construction and farmequipment such as bulldozers, combines, dumptrucks, backhoes, and scrapers. The company isorganized into a number of departments includingmarketing, product management, purchasing,human resources, manufacturing, and testing. Productand component designs are initiated by theproduct management group in response to competition,feedback from customers, or from engineers.The design process can be characterized as consistingof a number of phases starting with the productconcept, terminating with production, and anumber of reviews in between. The product/componentconcept is developed by the product managementgroup and is handed over to the teamwhich handles the design. The team, headed by aproject manager, is cross-functional in nature, consistingof representatives from each department.The concept document lists detailed specificationsof the component such as required horsepower,weight, maximum speed, blade capacity, angle,depth, and maximum capacity.6 The name is disguised. The case has been constructed by interviewingemployees who wished to remain anonymous.Knowledge and Process ManagementDepending on the level of experience, the task isassigned to an engineer within the team. The projectnow enters the alpha phase, where the emphasisis on the feasibility of the design. The design task isassigned to an engineer who develops two to threealternative designs for the component. The conceptis developed progressively, with weekly intra-teamreviews. The engineer develops two or three crudedesigns, specifying the general parameters of thecomponent, such as shape, size, material etc. Inthe process, the engineer consults previous designs,results of past simulations, etc. He/she receivesfeedback from team members regarding manufacturability,materials, etc. For instance, an engineermay be tasked to develop a door for a combinethat opens 90 . A materials specialist within thegroup may suggest a special type of hinge forthe door and a manufacturing technician may providefeedback regarding the area required forwelding the hinge in place.When the initial designs are thus developed, aformal presentation is made to a core group consistingof the senior representatives from eachdepartment. The group reviews the designs basedon criteria such as cost, robustness, manufacturability,and vendor availability, and selects onefor further development. If all the designs areequally satisfactory, a PUGH analysis is carriedout. In this type of analysis, the functional specificationsof each of the competing designs is ratedagainst various criteria such as criticality of component,failure mode, noise, and machinability. Oncea design is thus selected, the project enters the betaphase. The emphasis in this phase is proof of concept.A prototype of the component is developedand tested using computer simulations as well asa variety of test beds. Noise, engine specifications,stresses, fatigue, reliability, etc. are tested on thisequipment. The results of the test are maintainedin a database. Depending on the results, modificationsto the shape, hardness, and reinforcements ofthe component are carried out. For instance, if inthe transmission the gears are noisy, the materialmay be changed, or if there is leakage, additionalseals are added.The project now enters the gamma and deltaphases. In the gamma phase, the design is almostproduction ready. The component is installed inthe equipment and customers are brought in forfeedback. Depending on the component, customerscould give comments regarding the operating characteristicsof the machinery, such as ease of starting,speed, noise, ground clearance, andmanouverability. Further modifications are madeif necessary. During the delta phase, manufacturingsystems are tested. The idea is to ensure that parts72 C. S. Amaravadi and I. Lee

Knowledge and Process Managementare fed in, in the right order, that operators have thecorrect instructions, they are trained correctly andthe product is assembled in the right quantityand quality. At this stage, the design is enteredinto production. Structure: The design process can be roughlycharacterized as initiation, alpha phase, betaphase, gamma phase, delta phase, and production.Inputs to the process include specifications,prior designs, and feedback; outputs are theresults of reviews and tests, various proposeddesigns, modifications, manufacturing, andassembly instructions. Personnel and coordination: Qualifications andprior experiences of designers. What is the bestway to conduct product reviews? How to providespecifications? How best to transfer thedesign into production? Performance/tools: Knowledge of the engineer. It isvery specific to the task at hand and can be characterizedas the tacit or functional knowledge ofthe employee. Critical to this dimension are theresults of previous designs, access to test databasesand prior experience of the engineer. Discourse: Record of design changes, alternativedesigns considered, results of tests and reviews,problems identified, and changes made toaddress these problems. Results: The final output of the process, the actualdesign, whether it was on schedule, the amountof time it took, etc. Quality and objectives: Rather than quality, it is theobjectives that will be relevant here, since thecomponents are not yet in production. The basicobjectives will be those specified in the conceptdocument, viz. blade angle, bucket capacity,horsepower, etc. Implications: How can the design process bespeeded up? What types of components are feasiblein general? Which components will requirevendor sourcing? Is vanadium steel available in3-foot sections?KNOWLEDGE DIMENSIONSAND PROCESS TYPESRather than simply view processes as operational,administrative, and managerial, we have classifedthem into engineering/design, manufacturing/service,financial/accounting, administrative, legal,and managerial processes. These processes dealwith fundamentally different sets of inputs, outputs,and constraints. Not all dimensions are relevantfor all processes. Those that are relevant fora particular process type will depend on whetherRESEARCH ARTICLEor not the process is well defined, critical, involvescoordination, and has well-defined (i.e., measurable)outputs. In manufacturing processes, forexample, the structure and outputs are welldefined. Therefore dimensions such as structure,personnel, performance, results, and quality arevery relevant, while the discourse dimension andimpacts dimensions are not very relevant. Foradministrative, financial, and legal processes, thestructure, personnel, performance, discourse, andresults dimensions are very pertinent. On the otherhand, for managerial and design processes, therelevant dimensions are the discourse, results,and impacts, while other dimensions are not sorelevant. In the following, we give additional examplesof OP and identify some of the relevant itemsof knowledge. (Please note that examples of thestructural dimension have already been given inTable 1.)Industrial process—performance dimensionIn an industrial process to strip hydrogen sulfide(H 2 S) from waste gases, consider the followingdescription which fits in with the performancedimension (Anonymous, 2001, p. 35): ‘In caseswhere flaring will not reduce H 2 S concentrationsufficiently to meet the emission limit, ...othertreatment methods must be used. These includeiron sponges (and other iron-based absorbents),chemical scrubbers and water scrubbers. A recentlycompleted research and development project...determined that an iron oxide-based adsorptionmedium—Media G2—could efficiently andcost-effectively remove H 2 S from biogas.’ Theextract describes how H 2 S can be removed fromwaste gases and is therefore an example of the performancedimension.Consumer lending process—personneldimensionA study of 10 financial institutions identified a numberof best practices in the lending process. Amongthem, the study found that ‘in addition to the delinquencyrate, focusing on the cure rate and loan portfoliowere the optimal ways to determine anindividual collector’s productivity’ (Leath, 1998,p. 38). As described in the case, the ‘cure rate’ is aneffective method of evaluating loan collectors andillustrates one aspect of the personnel dimension.Purchasing process—coordination dimensionThe following is an illustration of the coordinationdimension in the UK food service industry: ‘TheThe Dimensions of Process Knowledge 73

RESEARCH ARTICLEKnowledge and Process ManagementTable 4Process types and dimensionsProcess type Examples Relevant dimensionsEngineering/design Furnace set-up, boiler inspection, new Discourse, results, impacts, and implicationsproduct developmentManufacturing/service Manufacturing Nylon, assembling Structural, personnel/coordination,motherboardsperformance and tools, results, qualityFinancial/accounting Preparing financial statements, auditing Structural, performance and tools, results,impacts and implicationsAdministrative Hiring employees, buying equipment Structural, personnel/coordination,performance and tools, discourse, resultsLegal Issuing stock, preparing labor contract Structural, personnel/co-ordination,performance and tools, discourse, results,impacts, and implicationsManagerial Strategic planning, negotiating a Discourse, results, impacts, and implicationssupplier contractcatering review group meets every six weeks and isused as a vehicle to coordinate the operational fulfillmentof consumer requirements at site level.This review group consists of the marketing manager,...The review group acts as a forum for discussionbut also has the power to veto or ratify aproposal. This decision is made by consensus ...’(Mawson and Fearne, 1996, p. 39). The constitutionand operation of the group are illustrative of oneaspect of the coordination dimension, althougheffective interaction techniques would be a morevalued component.Customer service process—quality dimensionA major computer manufacturer uses the followingmeasures for monitoring its customer service process(adapted from Davenport and Beers, 1995):percentage of product returns (2%), percentage oforders delivered on time (99%), number of callsanswered per day (2400), number of calls abandoned(80), amount of waiting time for callers(5 minutes). The numbers in parentheses indicateacceptable values of these indicators and are illustrativeof the quality dimension.Engineering design—results dimensionThe following is a hypothetical example of theresults dimension: ‘A car manufacturer found thatengineering and development of a new model cost$1,000,000, with $250,000 spent on developmentand the rest on tooling. The process required 25engineers, 100 indirect employees and 3 years tocomplete.’ If this information were linked withsales of the car, there is an opportunity for evaluatingthe effectiveness of the process.These rudimentary examples bear out thehypothesis that process knowledge can have severaldimensions, of which some are salient in certaintypes of processes, and that tapping it can beof utility in communications, process design, training,etc. The framework has several implications,which are discussed next.DISCUSSION AND IMPLICATIONSWe have considered process dimensions from aknowledge management perspective, although otherperspectives have been presented in the literature.There has been an extensive body of literature consideringprocesses from an engineering, i.e. workflowautomation (Stohr and Zhao, 2001) and reengineeringstandpoint, i.e. BPR (business processre-engineering) (Davenport et al., 1996). In workflowautomation, the emphasis is on modeling thestructure of the process and automating it withsoftware. Issues such as activities, their constraints,dependencies, and authorizations are considered(Stohr and Zhao, 2001). In BPR, the objective is specificallyto achieve process improvements. There isa special emphasis on identifying critical processes,developing measures, assessing their performance,making improvements, and assessing the costs andbenefits (Gardner, 2001; Lientz and Rea, 1998).Despite sharing the same objectives (processimprovement) and overlap in the information analyzed(in the structural and quality dimensions),these perspectives are not intended to tap processknowledge as we have attempted to do and besideslack holistic approaches to it.In attempting to provide a KM perspective, wehave avoided labels such as outputs (subsumedby structure and results), costs (also subsumed byresults), productivity (same), work-in-progress(not considered), and status (not considered) which74 C. S. Amaravadi and I. Lee

Knowledge and Process Managementcould potentially communicate a data orientation.We have also not considered process evaluation(subsumed by implications) and functional knowledge(subsumed by performance knowledge).Instead, we simply focused on the process learningcycle and used labels appropriate to this. Whetherthe labels are justified and whether or not they areadequate is difficult to validate in an empiricalsense because of the qualitative nature of the framework.They can, however, be refined experientiallyby being applied in various organizations,for various types of processes.The framework provides a preliminary foundationfor organizations in tapping knowledgeresources that are in the process form. In order togo about acquiring the knowledge, a more detailedcharacterization as it pertains to each dimension isrequired. (This has been carried out to some extentin the paper, although not formally). Whether ornot such a characterization can be carried outwith the KM perspective intact is debatable. Thestructural dimension has been characterized asinputs, steps, outputs, and constraints, but thishas little value except in the fully interconnectedform, because this is the nature of knowledge. Considerthe following description of an assembly process(Garg, 1999, p. 419): ‘Board 1 and Board 2 areassembled along with other parts into modules oftype A. Type B modules are manufactured at a differentsite. The Marry Station ...loads special softwarethat enables type A and type B modules towork together. However, by redesigning thesemodules, this operation can be eliminated altogether.’Clearly, structural knowledge is intertwinedwith the inputs, outputs, and operations.Communicating this knowledge is not possiblewithout the use of process charts. Even if it wereuseful, it may not constitute knowledge to a productionmanager, who might be more interestedin aggregate characteristics such as throughputsand capacities. As another example, the discoursedimension can be characterized as issues (can thesebe clearly distinguished?), actors, viewpoints, andtime periods. This perspective can yield informationon what a particular actor may have said at aparticular time, but not necessarily his/her motivationor objectives, which have to be evaluatedbased on the entire discourse. Utility aside, eachof the dimensions is also complex and interrelated.For instance, there is overlap between the structural,quality, and performance dimensions becauseincreasing quality will require changes to the toolingand the way activities are carried out. Similarly,the quality dimension can encompass measures foreach activity of the process, which could vary withthe type of product. A detailed characterization ofprocess knowledge could potentially be problematic,unless the context of the entire processwere somehow preserved and individual items ofknowledge were presented within that context(see also APQC, 1997). This responsibility restswith designers of KM systems.The design of effective KM systems is contingentto a large degree on the existence of effective methodologiesand tools. Methodologies are techniquesfor modeling process knowledge, and can encompassPetri nets, discourse maps, cause maps, etc.In the KM literature, tools are defined as thosewhich support one or more phases of the KM cycle(Tyndale, 2002; Wensley and Verwijk-O’Sullivan,2000). There are several dozen generic tools suitablefor this purpose, a simple example being atool to organize documents. However, we use toolsin a restricted sense to mean those which supportthe acquisition and codification of knowledge. Generallythese tools support advanced techniques suchas ontologies and knowledge maps rather than simplyusing rules.It is necessary to assess and, if needed, developsuch tools and methodologies to support the KMprocess. It is expected that there will be well-developedtools for the structural and discourse dimensions(see Stohr and Zhao, 2001; Amaravadi, 2004),but for other dimensions techniques from artificialintelligence (see Amaravadi, 2001) may have to beutilized. It is also expected that the methodologieswill have to be tailored for each type of process.Indeed, the development of effective knowledgemanagement systems is one of the principalgoals of the KM community which can greatlybenefit from improved codification techniques.This is even more critical in the case of processknowledge.ACKNOWLEDGEMENTSWe gratefully acknowledge constructive commentsfrom Dr Mandeep Singh, Department of Marketing,Western Illinois University, as well as fromour anonymous reviewers, which have greatly contributedto the paper.REFERENCESRESEARCH ARTICLEAmaravadi CS. 2001. Engineering administrative knowledgefor extended office systems. In 2nd EuropeanConference on Knowledge Management, Bled, Slovenia,7–8 November.Amaravadi CS. 2004. Research issues in advanced officeinformation systems [available from author].The Dimensions of Process Knowledge 75

RESEARCH ARTICLEAnonymous. 2001. Process effectively strips hydrogensulfide gas from wastewater. Water Engineering andManagement 148(11): 35.APQC (American Productivity and Quality Center).1997. Using Information Technology to Support KnowledgeManagement. American Productivity and QualityCenter: Houson, TX.APQC (American Productivity and Quality Center).2004. Process Classification Framework. American Productivityand Quality Center, Houson, TX http: / [May 2004].Argote L. 1999. Organizational Learning: Creating, Retainingand Transferring Knowledge. Kluwer: Norwell, MA.Ball LJ, Evans JSBT, Dennis I. 1994. Cognitive processesin engineering design. Ergonomics 37(11): 1753–1786.Curran T, Keller G, Ladd A. 1997. SAP R/3 Business Blueprint:Understanding the Business Process Reference Model.Prentice-Hall: New York.Davenport T, Beers MC. 1995. Managing informationabout processes. Journal of Management InformationSystems 12(1): 57–80.Davenport T, Prusak L. 1998. Working Knowledge.Harvard Business School Press: Boston, MA.Davenport TH, Jarvenpaa SL, Beers MC. 1996. Improvingknowledge work processes. Sloan Management Review37(4): 53–65.Demarest M. 1997. Knowledge management: an introduction.Long Range Planning 30(3): 374–384.Gardner RA. 2001. Resolving the process paradox. QualityProgress 34(3): 51–59.Garg A. 1999. An application of designing products andprocesses for supply chain management. IIE Transactions31(5): 417–429.Garvin D. 1997. The processes of organization and management.Sloan Management Review 39(4): 33–50.Gerson EM, Star SL. 1986. Analyzing due process in theworkplace. In Third ACM SIGOIS Conference, Providence,RI, 6–8 October, pp. 70–78.Hackbarth G, Grover V. 1999. The knowledge repository:organizational memory information systems. InformationSystems Management 16(3): 21–35.Hewitt C. 1986. Offices are open systems. ACM Transactionson Office Information Systems 4(3): 271–287.Kim DH. 1993. The link between individual and organizationallearning. Sloan Management Review 35(1): 37–50.Leath JT. 1998. Study identifies best practices in core processesof consumer lending. Journal of Retail Banking20(1): 35–39.Leppanen A. 2001. Improving the mastery of work andthe development of the work process in paper production.Relations Industrielles 56(3): 579–605.Knowledge and Process ManagementLesser E, Wells L. 1999. Managing organizationalknowledge in a revenue agency: the experience ofWashington DC. Government Finance Review 15(2):21–24.Lientz BP, Rea KP. 1998. Business Process Improvement.Harcourt Brace.Lubit R. 2001. Tacit knowledge and knowledge management:the keys to sustainable competitive advantage.Organizational Dynamics 29(4): 164–178.Malone TW, Crowston K, Lee J et al. 1999. Tools forinventing organizations: toward a handbook oforganizational processes. Management Science 45(3):425–443.Martiny M. 1998. Knowledge management at HPConsulting. Organizational Dynamics 27(2): 71–77.Mawson E, Fearne A. 1996. Purchasing strategies anddecision-making processes in the food service industry:a case study of UK restaurant chains. Supply ChainManagement 341(3): 34–41.Pearlson K. 2000. Managing and Using Information Systems.Wiley: New York.Pentland BT. 1995. Grammatical models of organizationalprocesses. Organization Science 6(5): 541–556.Ruggles R. 1998. The state of the notion: knowledge managementin practice. California Management Review,Special Issue 40(3): 80–89.Stohr EA, Zhao JL. 2001. Workflow automation: overviewand research issues. Information Systems Frontiers 3(3):281–296.Tyndale P. 2002. A taxonomy of knowledge managementsoftware tools: origins and applications. Evaluation andProgram Planning 25(2): 183–190.Van der Bent J, Paauwe J, Williams R. 1999. Organizationallearning: an exploration of organizational memoryand its role in organizational change process.Journal of Organizational Change Management 12(5):377–405.Walsh JP, Ungson GR. 1991. Organizational memory.Academy of Management Review 16(1): 57–91.WFMC (Workflow Management Coalition). 1999.Terminology and glossary, Document # WFMC-TC-1011.[December 2004].Wensley AKP, Verwijk-O’Sullivan A. 2000. Tools forknowledge management. In Knowledge Horizons,Despres C, Chauvel D (eds). Butterworth Heinemann:Oxford; 113–129.Wiltzer KH. 2000. German engineering firm developsnew Nylon production processes. International FiberJournal 15(3): 85–87.76 C. S. Amaravadi and I. Lee

More magazines by this user
Similar magazines