2. The vision of the digital factoryThe vision of the digital factory could be as follows:All computer-aided tools necessary for the plan-ning of new products and production plants as well asfor the operation of the factories are networkedthrough a central database. New structures in theproduct development and manufacturing processesensure that the requirement for simultaneous engi-neering is satisfied. Thus, the entire factory issimulated on the computer as a continuous andconsistent virtual-reality model (VR model), whichcan then be applied, without interruption, all the wayfrom the product idea to the final dismantling of theproduction plants and buildings. An automatic datamanagement system ensures that changes result in theupdating of the data in all company departmentsconcerned after their release. Access to all necessaryinformation is possible on a permanent basis. The datacan be exchanged between systems without anyconversion whatsoever, since the open structure of thesoftware tools allows the docking of new company-specific modules and new tools. Efficient VR systemspermit high-end visualization of all situations at anytime; hence, interdisciplinary cooperation amongvarious experts is possible all the way from theproduct design to the inspection of the new ormodified factory. The simulation techniques have been continuouslyimproved. As a result, the phases of physical producttesting can also be minimized; consequently, thenumber of product prototypes also decreases consider-ably. For production plants, the try-out is performedentirely with the aid of VR-supported simulations;hence, there is no need to construct physical pilot plants.The construction of new production plants can bepermanently monitored by means of laser scanning.The VR models generated from the scans of the hallsand plants can be compared with the planning models;errors in implementation can thus be detectedimmediately. Remedial measures or planning adapta-tions can therefore be implemented directly.Pilot production is no longer necessary, since avirtual confidence test is also performed here.Consequently, the start-up curves for production arealso very steep, since only a few errors still occur in thesystems or their logistic interconnections. During theproduction phase, material-flow simulators monitorand control the factory operation and provide supportfor job scheduling. If major malfunctions occur,emergency scenarios are quickly generated andsimulated; thus, clear-cut procedural instructionscan be provided for an efficient reaction. Conversionand optimizing processes can likewise be simulatedwith the computer model; hence, optimal conditionscan also be created for reorganization, conversionphases, and start of operation. Changes in the layoutsresult in an automatic adaptation of the associatedmaterial-flow models and can be immediatelycompared with simulation results and appraisedcorrespondingly. With an extension of the simulationfunctions, permanent cost monitoring is possible. Inthis context, the digital factory also functions as a toolfor controlling and for more extensive economicconsiderations.During the developmental phases, the digitalfactory thus operates virtually in the same way asthe real factory should function after implementation.Furthermore, it is permanently coupled with the actualproduction process after the realization of the plansand can thus be employed for monitoring, controlling,and constantly improving this process. In Fig. 1, thevision of the digital factory is represented graphically.3. Problems along the way to the digital factoryAn example from the history of industrial devel-opment during the past decades very clearly illustratesthe far-reaching consequences, which can result from  the increased application of computer-aided tools.
Theintroduction of ‘computer-aided design’ (CAD) toolscan be viewed as a kind of miniature revolution.Because of general scepticism and inadequate userorientation of some CAD programs, the completeimplementation of such tools in industry as a wholerequired decades. Initially, it was simply not possibleto substitute such tools for the knowledge and skill ofengineering draughtsmen. Awell-versed draughtsmanwas capable of achieving a usable result on hisdrawing board considerably faster than was possiblewith the use of a computer. Furthermore, the efficiencyof computers and plotters was, in many cases, still notsufficient for attaining the required quality. Never-theless, the advantages resulting from the considerablesimplifications in the management of change and fromthe performance of time-consuming routine activitiesfinally paved theway for widespread acceptance of thetools. To a steadily increasing extent, the preparationof detailed drawings and plans was assumed by thedesigners themselves; this development resulted indecisive structural changes in some functionaldepartments in the companies, as well as in changesin the methods of design and construction. Thedistribution of work over various stages of planningdevelopment thus became possible, too.However, disadvantages also resulted from theintroduction of CAD tools. Over the years, manyengineering draughtsmen had accumulated highlyspecialized know-how in the course of their activities.Practicable detail solutions were often the work ofdraughtsmen, and the avoidance of certain errors wasrendered possible by their work. As the drafting depart-ments were phased out, the loss of valuable know-howwas thus accepted as an inevitable consequence.From this example, it is evident that the introduc-tion of a new system decisively affects structures andprocedures. In this context, the implementation of thedigital factory can be expected to cause considerablechanges in previous production processes. Computer-aided monitoring and control of the factory operationwill also be affected.Essentially, four major problem fields result fromthe implementation of the digital factory. Theseproblem fields are considered in more detail in thefollowing and are illustrated in Fig. 2. 3.1. NetworkingThe fundamental requirement of comprehensivenetworking among all software tools involved in theproduction processes and in factory operation ishindered by their predominantly proprietary design.As is the case of the CIM concept, native formats giverise to a multiplicity of interfaces, which rendercooperation among the various tools difficult or evenimpossible. The expenditure for the conversion of therequired data is still considered to be too high atpresent and is therefore often regarded as unaccep-table.3.2. Version, knowledge, and data managementAs the efficiency of the software and hardwareincreases, the working habits of the user also change.Especially the increase in memory capacity results inimmense volumes of stored data. Hence, the memoryrequirement has increased dramatically during recentyears, and a veritable flood of data must therefore bemanaged. This demand is in part due to the steadilyincreasing use of new tools in product development,factory planning, and factory operation. On the otherhand, however, the modern user is intensivelyconcerned with data back-up. Consequently, a largenumber of data-record versions are stored in thecourse of a project, in order to ensure the continuedavailability of the data in the event that they are neededagain. Furthermore, identical data are often storedredundantly on various servers and systems.A further challenge results from the necessity ofmanaging and securing a company’s specific know-how in a manner appropriate for preventing loss.Workflows and experience must be recorded insuch a way that they are available and useful to allemployees.Obsolete data-record versions and insufficientinformation flows constitute sources of error whichshould not be underestimated. Since the developmentand planning process is characterized by a multiplicityof iteration loops, adaptations are absolutely necessaryin all departments of the company and in all phases ofproject execution. Particularly in the digital factory,special demands are imposed on the maintenance ofdata and the management of versions in this respect.It must be ensured that up-to-date, subject-specific,relevant data are continuously and immediatelyavailable to all those concerned, especially in theevent of changes. The large quantities of data and theequally large number of different data formatsconstitute one of the main problems associated withthe digital factory.3.3. Processes and structuresIt is assumed that the existing structural organiza-tion of company departments will change as the digitalfactory becomes more widespread. Fewer physicalprototypes will be required, and various specializeddepartments will be combined to an increasing extent.A redistribution of functions and responsibilities willbe necessary for creating integrated and efficientdesign processes. Moreover, the concerned personnelwill be confronted with new planning procedures. Theconventional phase models of project work mustlikewise be eliminated, since a clear-cut delimitationof inpidual steps no longer makes sense if thepotential of the digital factory is to be fully utilized. Inthis context, simultaneous engineering will become anecessity, and its possibilities will be utilized to thelimits.At present, only speculation is possible in connec-tion with the actual structures of the hierarchical andsequential organizations within companies in thefuture. In any event, it is certain that decisive changeswill occur. As already pointed out in a publication [3],the essential supporting pillars for the digital factorymay already be standing by 2005. However, theconversion process itself will require much more time,more than was required for the introduction of CADsystems. Furthermore, considerable research andinnovative solutions will be necessary.3.4. CommitmentAt present, the digital factory is still a project forlarge-scale enterprises, especially for OEM’s in theautomotive industry. Large amounts of capital arebeing made available for making the vision come true.However, questions concerning the necessary level ofimplementation are also receiving increased attention.Conversion of the incurred costs to yield a financialreturn flow is not yet possible. Of course, many of thesoftware tools necessary for processing inpidual partial aspects of product and production developmentare already available and have also been tried andproved in applications. However, the availableresources are not yet sufficient to allow for a rapidand useful application of these tools, especially insmall and medium-sized enterprises. Consequently,not only internal structures are subject to change; inparticular, cooperation among companies must also beadapted accordingly. This matter is currently beingconsidered, too, and the objective is to achievecontinuity at least up to the 2nd tier suppliers.In the course of these deliberations, however, onequestion is seldom considered: are the SME, who arevery likely to be among the 1st or 2nd tier suppliers,really capable of adequately implementing the digitalfactory? These smaller companies usually do not havelarge planning departments of their own for dealingwith production requirements, and they usually cannotafford the corresponding software tools. Conse-quently, it will be difficult for them to utilize thepotential offered by the digital factory [1]. For manysmall companies, investments of this kind are hardlyfeasible, since they are not economically justifiable aslong as the price of the necessary software remains sohigh, and current expenses can hardly be borne bysmall companies. Initial approaches toward possiblesolutions have already been attempted [5], but havenot yet found widespread acceptance. Major reasonsfor this reluctance certainly include the lack ofinformation and the absence of appropriate structureson the part of the KMU. Nevertheless, the digitalfactory approach is not merely an inevitable necessity;it also offers a chance for some medium-sizedsuppliers. In this case, too, considerable potentialsavings could be achieved by implementing theappropriate structural measures and with supportfrom external service companies.4. The reality of the digital factoryThe vision and the associated problems have beendescribed. However, what is the current develop-mental status of the digital factory? In this context, itshould first be mentioned that the level of implemen-tation already achieved by some companies cannotbe determined exactly. Hence, the following dis-cussion should be regarded as a summary of theauthors’ experience and an analysis of the publishedinformation.At least one thing is certain, however, largeinvestments in the digital factory have already beenrealized, especially in the automotive industry and bymany 1st tier suppliers. Furthermore, it cannot bedenied that companies have already achieved successin certain partial fields and in pilot projects. Anexample is the nearly uninterrupted application ofvarious planning tools by automobilemanufacturers invehicle body construction. Product data integration,robot simulation, and the associated off-line program-ming have already become established there. Twolarge software suppliers currently constitute thesupporting columns of the digital factory. As indicatedby current knowledge, however, it is not necessarilyprobable that a software platform can cover all facetsof the digital factory. Previous cooperation among thesuppliers of tools, which may be applicable within thescope of the digital factory is by no means adequate,and the demands imposed on a central, standardizeddata management system can hardly be satisfied atpresent. For small and medium-sized enterprises,many of the tools are still much too expensive, or costrecovery from applications is not possible. None-theless, some of these companies are alreadyrequesting delivery of certain proofs or data in digitalform to match their customers’ respective digitalfactory.An increasing number of system suppliers alsooffer services in conjunction with the digital factory.Since the various computer-aided tools in thedevelopmental, design, and planning stages are sonumerous, however, these suppliers must also come togrips with a multiplicity of interfaces. Furthermore,they are forced to keep almost all software toolsavailable, in order to be able to satisfy the needs of asmany customers as possible. This requirement alsoposes the question of economy and will furtheraggravate competition.Hence, it must be concluded that inherentlynecessary, standardized networking of the tools ishardly feasible, since the structures of many tools donot allow for docking with other modules orapplications, or at least render an adaptation difficult.Furthermore, the problems of version and datamanagement have not yet been solved. These factorscould further increase the multiplicity of insular software solutions in the various company depart-ments. Regardless of which decision is reached infavour of a particular general supplier of software forthe digital factory, certain advantages as well asdisadvantages will always result from the choice. Forinstance, it cannot always be assumed that allproblems associated with the interfaces betweensupplier-specific software tools have been solved.Since the existing company structures tend to reflectconventional patterns of thought and work, and sincethe new process and method orientation is not yetpracticed in a comprehensive manner, only a limitedportion of the potential savings can be achieved withthe insular IT solutions. The real possibilities offeredby the digital factory are accessible only through anappropriate networking of all tools in combinationwith a restructuring of the processes and hierarchicalorganization of the companies. Reconsideration andredistribution of competencies and responsibilities willalso be necessary.Furthermore, a considerable untapped potential isstill available from the computer tools themselves.The user interfaces of the tools and the user guidancemust become considerably more intuitive. After all,the user’s creativity should not be subordinated to theformalities of the tools; in particular, ideas must beimplemented and tested quickly and simply with theuse of the computer.5. Small steps toward success—an example ofresearch at IMABThe following project description should provideinsight into the fundamental research in progress atIMAB on the subject of the digital factory.In order to allow for interdisciplinary discussionsamong the project participants from different fields ofspecialization, the concept of the virtual communica-tion platform is frequently employed in conjunctionwith the digital factory. Above all, this term designatesa virtual-reality-oriented tool, which is employed toallow for a more effective visualization of variousplanning results. For instance, the results of the designprocess can thus be considered together with thefactory planners’ data and the logistician’s concepts.Joint planning meetings before and during VRprojection thus allow for more efficient mutualagreement on the work and support the continuingdevelopment activities on the basis of results whichare already available. An example is the so-calleddigitalmock-up, which permits the virtual assembly ofinpidual components and subassemblies. A furtherfield of research is the networking of results frommaterial-flow simulation and layout planning. In thiscontext, the functions of a material-flow simulatorhave been integrated into a VR model developed atIMAB [2].5.1. ObjectivesFor the new manufacturing technology based onhigh-frequency welding in the sheet-metal industry, ageneral survey already indicates the possible range ofseries applications in future production, even at a veryearly stage of development. The short process time forthe welding of sheet-metal profile components by thismethod had suggested the possibility of applying thistechnique to the manufacture of modular-designcomponents. The associated requirements on thestructure of a production plant should be investigated,especially from a logistics point of view at this earlystage of process development.5.2. ProcedureIn order to find a realistic approach to the problemsposed by the new manufacturing process, a fictitioussubassembly was designed; this subassembly con-sisted of three inpidual components and could beproduced in 10 different versions by high-frequencywelding. From the structure of the subassembly, thenecessary manufacturing facilities were then derivedand combined to form a production cell in apreliminary draught. For this purpose, two of thethree components had to be manufactured directly inthe production cell. For one of these two components,five versions of different shape and material wereenvisaged. Since a metal-shaping press was to be usedfor its manufacture, a component buffer had to beincluded in the system to ensure continuity of supplyfor production. In Fig. 3, the virtual prototype for theprocess is illustrated for high-frequency welding.On the basis of the preliminary layout draught forthe production cell, the construction of a model formaterial-flow simulation was attempted, in order to  test the functioning and efficiency of the system. Uponparameterization of the simulation model, however, anexact determination of the times – almost by memory– proved to be impossible formost of the manipulationand clamping operations as well as for the entireprocess sequence in the high-frequency weldingfacility. For this reason, and also in view of thetechnical possibilities offered by the VR laboratory atIMAB, it was decided to construct a VR model. Thepurpose of this model was to visualize the kinematicsof the welding facility as well as all operations, inorder to allow conclusions to be reached regarding theactual process and operation times. From the basicdesign and conception of the welding facility, ananimated VR model of the production cell wasconstructed within a period of 5 months. For thispurpose, the system layout was structured and furtherdeveloped by an interdisciplinary project team duringseveral work sessions in the VR laboratory. Thediscussions among the specialists were supported in adecisive way by the three-dimensional models and thestereoscopic, large-screen projection; as a result, thevirtual reality assumed the function of a communica-tion platform. 虚拟现实的数字化工厂英文文献和中文翻译(2):http://www.751com.cn/fanyi/lunwen_44920.html