Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tppc20
Download by: [National Inst of Intellectual Property] Date: 23 October 2015, At: 06:27
ISSN: 0953-7287 (Print) 1366-5871 (Online) Journal homepage: http://www.tandfonline.com/loi/tppc20
How engineering data management and system
support the main process-oriented functions of a
large-scale project
Ari-Pekka Hameri & Juho Nikkola
To cite this article: Ari-Pekka Hameri & Juho Nikkola (1999) How engineering data
management and system support the main process-oriented functions of a large-scale project, Production Planning & Control, 10:5, 404-413, DOI: 10.1080/095372899232920
To link to this article: http://dx.doi.org/10.1080/095372899232920
Published online: 15 Nov 2010.
Submit your article to this journal
Article views: 45
View related articles
PRODUCTION PLANNING & CONTROL, 1999, VOL. 10, NO. 5, 404–413
How engineering data management and system
support the main process-oriented functions of a
large-scale project
ARI-PEKKA HAMERI and JUHO NIKKOLA
Keywords con guration management, engineering/product data management, large-scale projects, project management, World Wide Web
Abstract.This paper studies the bene ts of establishing con- guration management procedures and using an engineering data management systems ( EDMS) in a large-scale project by dividing the overall eå ort into successive processes. Through the identi cation of the main functional groups who will use the EDMS, the paper outlines the basic motivations and services provided by such a system to each process during the life-cycle of the project. The implications of strict con guration manage-ment on the daily operation of each functional user group are also discussed. The main argument of the paper is that each and every user of the EDMS must act in compliance with the con- guration management procedures to guarantee the overall bene ts from such a system. In truly global projects, special emphasis is placed on the management of distributed
opera-tions, which sets the exible and universal access to all engin-eering and project related documents into a crucial role in the overall success of the project. As an example of the given chal-lenge, this paper documents a World Wide Web-based system to manage documents and interface in-house document vaults and commercial EDM systems. The underlying environment of the case is that of CERN and the ongoing, decade-long, Large Hadron Collider ( LHC) project.
1. Introduction
The fundamental role of a con guration management is to control the storage and distribution of technical documentation, and this way support the multiple pro-cesses related to the product’s design, manufacturing, assembly, inspection, testing and maintenance, i.e. its whole life-cycle. In a large-scale project with globally Authors: Ari-Pekka Hameri and Juho Nikkola, Helsinki Institute of Physics, CERN EST-Division, CH-1211, Geneva 23, Switzerland.
Ari-PekkaHamerireceived his degrees of Master of Science and Licentiate of Technology at the Helsinki University of Technology ( HUT) , both concerning production management. He was conferred with the degree of Doctor of Technology from HUT in 1993 for his studies related to innovations and their technology impact on manufacturing companies. For two years he was the director of the Institute of Industrial Automation at the HUT, and currently acts as the director of technology program within the Helsinki Institute of Physics. He has been involved with several EC-funded and other international research projects dealing with production management and logis-tics. Currently he is associated with CERN and the con guration and communication manage-ment issues related to the construction of the new accelerator.
JuhoNikkolareceived his Master of Science degree in Industrial Management at the Helsinki University of Technology in 1990. Prior to joining the Helsinki Institute of Physics to work at CERN he held diå erent positions at the Institute of Industrial Automation at HUT, ending up in the position of vice director. At CERN, he has been involved in EDMS prototype development, selection of a commercial product, pilot installations and implementation. Currently he is respon-sible for developing WWW-based support for con guration and engineering data management of the new accelerator.
0953-7287 /99 $12.00 Ñ 1999 Taylor & Francis Ltd.
distributed design and production processes, the task to manage all product-related information is of immense importance for the overall project success. This necessi-tates the use of well-de ned con guration management processes and software tools capable of maintaining and controlling all the data and processes related to the devel-opment and implementation of the product. These issues are of current concern at CERN* where the design of the Large Hadron Collider ( LHC) is well advanced and oæ -cial plans for its implementation are being devised. By considering earlier accelerators built at CERN, the LHC speci cations will comprise about 300 000 drawings and technical notes, which are produced in almost 40 diå er-ent countries, as is the case with the particle experimer-ents. Thus, some kind of a product data management/engin-eering data management system ( PDM/EDMS) and related procedures must be established to keep this global intellectual eå ort on the right track.
Traditionally, the functions of con guration manage-ment have been divided into four main areas: Con guration Identi cation, Con guration Control, Status Accounting and Con guration Audits. This divi-sion provides only a partial support for analysing the data management needs for life-long support of product data, a need typical for large-scale projects. The applica-tion of engineering data management tools for project-based business has been limited, but currently these needs are driving PDM/EDM systems towards the traditional areas of MRP ( Manufacturing Resources Planning) and CMMS ( Computerized Maintenance Management Systems) . Thus, several activities are prevailing in a large-scale project, rst the general project management related activities concerning scheduling, progress control and budgeting must be carried out. If, as often is the case, the project involves a signi cant design phase, then con- guration management procedures must exist and be applied thoroughly. Finally, the manufacturing, as-sembly and installation activities require production management skills. One thing is certain, that each of these management activities requires document and change process control in order to be successful in their tasks. In the following, the main focus is on con guration management, yet with multiple implications to project and production management activities.
Con guration management involves the coordination, dissemination and validation of all the information related to product con guration changes. The ultimate goal is to maintain the coherence between the volumi-nous amount of technical speci cations and other data in order to issue and maintain diå erent versions of the prod-uct design during its development process. In order to manage this task, a team is required to mediate con- rmed information within the project organization. An EDMS will provide support for this diæcult task. The services provided to a project engineer by the con gura-tion team vary depending on the phase of the project. By dividing the project’s life-span into three main steps, i.e. conceptualization, execution ( including inspection and testing) and exploitation, the following table ( table 1) broadly summarizes the distinctive roles of con guration management and EDMS. The main underlying function of engineering data management is to provide the whole project organization with a coherent and reliable source of product information.
The rest of the paper is organized in the following way. First, before outlining the main processes and their needs of product data service, we brie y look at the basic indus-trial motivations for con guration management. The aim is to highlight how seriously industry is taking the con- guration management-related investments and what gains have been made from these investments. The pro-ject life-span is then divided into process phases to inde-pendently de ne the bene ts and services of each step of the project, followed by a review of the preliminary results from the pilot tests with the prototype system being developed and used at CERN. Finally, conclusions are drawn in the form of normative suggestions on how to establish parallel procedures and information systems that support each other through the life-cycle of the project.
2. Motivation from an industrial point of view
Industrial companies have realized the huge potential hidden in the control of design and engineering changes. The annual markets around EDMSs are estimated to boom from 1994’s $500 million with an average 30% annual growth to around $2.0 billion in 2000 ( CIMdata 1994, 1997) . It has been predicted that doc-umentation management will be among the fastest grow-ing software niches after the network applications by the end of this century. The reasons for this are obvious. Manufacturing enterprises have mostly focused their development eå orts on operations management accord-ing to main prevailaccord-ing manufacturaccord-ing hypes, e.g. business process re-engineering and lean manufacturing. The next potential area of development emerges from eæcient
* CERN, the European Laboratory for Particle Physics, has its head-quarters in Geneva. At present, its Member States, Austria, Belgium, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Italy, The Netherlands, Norway, Poland, Portugal, the Slovak Republic, Spain, Sweden, Switzerland and the United Kingdom. Israel, the Russian Federation, Turkey, Yugoslavia ( status suspended after the UN embargo, June 1992) , the European Commission and Unesco have observer status.
information management. The ultimate goals of both of these drives remain the same: reduce time to market, shorten design review cycle, reduce manufacturing lead times, increase productivity, reduce rework, etc. Another necessary prerequisite is that all documents will even-tually be produced and handled in electronic form.
Although it is diæcult to obtain exact quantitative data on the net bene ts and investments on EDMS and con guration management, some facts, gures, results and management actions may be found to be publicly available:
Several US based companies have started to estab-lish electronic archives for their technical documen-tation by scanning thousands of technical data items into a database ( Puttre 1992) .
A company producing large industrial production units indicated that on average half of their design and engineering eå ort is related to searching for technical documentation ( Nihtila¨ 1996) .
One industrial facility producer estimated that they have 4 million technical drawings, and when a new order comes in they have 80% probability of having the required drawings ready, yet the odds to nd them are 4% ( Sulonen 1995) .
A global electronic company reduced the change order cycle from 33 days to less than 10, halved the time needed to devise technical manuals and eliminated micro- che through the introduction of a product data management system with the objec-tive of paperless documentation management ( Frederick 1995) .
A multinational computer and electronic company reduced its staå required to manage documentationload by 35% , plummeted storage cost/year/docu-ment from 50$ to 5$ and improved signi cantly the document handling eæciency during a ve year EDM project during which its own EDMS was developed ( McCrea and D’Agostino 1995) .
A leading microprocessor manufacturer developed with the eå ort of seven engineering years global, standard and PC-based EDM procedures with ven-dor access; the results are eæcient document hand-ling, paperless document processes and up-to-date information ( Hargrave 1995) .The above examples illustrate well the potential hidden in eæcient con guration management and in logically centralized storing of all information. Predominantly, the results have been attained through the introduction of new managerial principles supported by an EDMS to maintain, control and archive all the design and speci cation changes. Multiple sources report that investments on EDM and supporting systems have pay-back times from one to three years. One should be cautious here, as these times are calculated from the point on where the systems have been widely used by the organization. Experience seems to show that major com-panies have been testing several commercial systems before turning into one vendor or even returning back to in-house built system. From any organization’s point of view, the needs and motivations for con guration management and the related EDMS originate from the following facts:
design and manufacturing operations are becoming ever more distributed through increased out-sour-cing and tending to focus on core competencies;406 A.-P. Hameri and J . N ikkola
Table 1. The value-adding functions and services of con guration management and EDMS for a large-scale and distributed project. The supportive actions of con guration management and EDMS
Conceptualization Execution Exploitation
Con guration
translates customer requirements
ensures product’s manufacturability
supports operational andmanagement into technical language
disseminates technical information maintenance activities of the system
manages version handling between vendors and project teams
provides a base of learning for the
maintains product coherence
controls speci cation and design future
controls speci cation and design changes
summarizes speci cation and designchanges changes
Engineering data
archives design history
provides continuous source of
provides data for an aftermath and management
provides user interface to up-to-date product information learning from the design process systems all product data
supports technical communication
helps tracing solutions to mainteance
maintains design process status between project collaborators from and operational problems
supports dissemination of bids to project closuretechnical information
Main
!
support and coordination!
mediation between customer,!
maintain operational period and contriubtion of engineering work/ production and design establishes learning process of EDMS information/data ow!
quality of manufacturing!
lifetime support!
quality of design
increased product complexity implies more docu-ments to be produced, distributed and managed;
most of the documents needed are produced in elec-tronic format, which makes their computerized management and distribution possible;
personnel of the organization keep changing, thus knowledge must be stored somewhere to secure the product’s and project’s whole life-cycle;
multi-project environments, with several open deliveries or projects at the same time, enforce the use of formalized procedures, which in a distributed operating environment should be supported by soft-ware applications.These requirements ask for a exible project organiz-ation with a capability to evolve and maintain an ade-quate level of expertise through the conceptualization, design, manufacturing, assembly, installation and opera-tional phases of the project ( Hameri 1995) . To manage this, the existing expertise of the organization must be carefully mapped and allocated along with the on-going projects to correspond to the technological require-ments and other project objectives. In addition, the expertise and related technical documentation should be controlled and stored to maintain the con guration coherence and their availability as the organization changes and diminishes its resources during the life-cycle. Below, we concentrate on the bene ts of con gura-tion management to the project organizagura-tion and its individuals.
3. Process-oriented functions—the user point of view
Individuals tend to execute many roles during the pro-gress of the project, e.g. an engineer may be responsible for designing, manufacturing and even installing a
cer-tain component and, thus, the services needed depend on which phase the work is progressing. To avoid misunder-standings by classifying the users according to their oæ -cial function, we proceed by describing the user point of view in terms of the processes or functions needed to nish the project. In the Introduction, the whole project was divided into three main phases, i.e. conceptualiza-tion, execution and exploitation. These phases are decomposed further. Figure 1 summarizes the further division of the project into consecutive processes.
The main services and bene ts of an EDM and related systems ( together to be referred to as EDM procedures) are outlined below from the point of view of each process of a large-scale project. The perspective is that of the LHC project, which can be compared to any largescale and geographically distributed projects with signi -cant technological challenges requiring several years’ eå ort and global collaboration. The LHC project has reached its technical baseline and, in general, is currently entering the execution phase, yet some parts of the pro-ject are already completing their assembly and installa-tion phases, e.g. the civil engineering works and other infrastructure-related activities.
3.1.Project management/administration/nance
Planning, scheduling and controlof the project resources and ow of work is the fundamental activity of pro-ject management. Decisions concerning resource allocation may be supported by following the amount of engineering changes and how they have evolved in each major task of the project work struc-ture. Once the project is progressing, the main task of project management is to keep the project on schedule and to solve problems as they emerge. Detailed reporting of the design changes and otherProject management/administration/finance
Þ planning, scheduling & control
Þ risk management
Þ supplier integration
multiple interaction and feed-back
Conceptualisation Execution Exploitation
Þ abstraction
Þ conceptualisation
Þ technical design & engineering
Þ technical documentation Þ manufacturing Þ quality management Þ assembly Þ installation Þ operation Þ maintenance Þ learning Þ system demolition Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð Ð time Ð Ð >
Figure 1. The main process-oriented functions of each task to accomplish the LHC system.
problems retarding progress are mediated almost on-line to the project management via eæcient EDM procedures.
Risk management and anticipation of possible near future technological pitfalls is a diæcult, yet import-ant, task of the project management. The preven-tive problem-solving activities bene t greatly from up-to-date information on how technical develop-ment is progressing and what problems have been spotted. In global projects, this information also has the possibility to follow how communication has evolved around certain technical documents and who has required them. This feature of electronic communication around EDM procedures provides the management with an extension to normal pro-gress reporting to know how the work has actually been progressing. Tracking of the communication behaviour gives the management an extra facet to tackle and anticipate technological risks.
Supplier integrationand logistics is a mandatory activ-ity for any large-scale project. The operational system produced by the project is the ultimate sum of multiple contributions from a huge variety of public, academic and industrial partners, most of them needing a exible and reliable access to the diverse technical documents de ning the system. This requirement of integrating the suppliers with the EDM procedures is vital to obtain the full com-mitment of the suppliers. In addition, it establishes the means to control suppliers and to communicate with them on a technical level. Also, issues concern-ing data security are of great importance when external communication is concerned.3.2.Conceptualization
Abstractionis a process, typical to large-scale scienti- c projects, during which the ultimate system, e.g. the LHC, is disassociated from the overall realm of thoughts around the subject matter. With LHC, this process results in the lattice model of the system with its capabilities, objectives and features termed in the language of physicists. The lattice model already de nes the main technological features of the system and acts as the initial model for the plan-ning of the overall system. As the project proceeds, the lattice serves merely as a view among the others in the system. The EDM procedures should take the lattice model as the rst model of the system, and later on, as the work proceeds, link it with other views of the system. The diå erent views of the system refer to the fact that the EDMS must be exible enough to provide a working environment customized to serve the peculiar needs of the diå er-ent processes and technologies involved with the project.
Conceptualization processes and concretizes the lattice model or system abstraction into a conception of the detailed technical needs with key parameters and their values. At CERN, the conceptual model at a moment in time can be seen to be the so-called par-ameter book describing the technical speci cation of the system, similar to design envelopes in many industrial applications. Yet, this book is already obsolete when printed and the EDMS would pro-vide a dynamic book. The conceptual model estab-lishes the fundamental structure for the technical document handling. This underlying structure acts as the skeleton of the product breakdown structure. EDM procedures are used to maintain the concep-tual model in coherence with the original lattice model. At its best, the EDMS serves as a commu-nication medium during the process of establishing the model. Ultimately, the EDMS will provide all project collaborators with easy access to the up-to-date conceptual model of the system.
T echnical design and engineeringprocesses are the prime bene ciaries of the EDM procedures. Design ma-terializes the conceptual model into technical details with their planned interaction with each other to meet the original functional requirements. Engineering activities further adjust the design into exact technical speci cations with material, manufacturing and assembly instructions. This activity also produces various technical notes and test reports, which are vital to the coming activities. EDM procedures provide the designers and engin-eers with an intuitive and easy access to the actual technical information, i.e. documents and controls their coherence with each other and compliance with the original plans. The management pro-cedures controlling the design and speci cation changes bene t from the EDMS as media to archive, distribute and control the design status of the system.
T echnical documentationprocess during the whole con-ceptualization phase generates a huge amount of technical notes, test reports, drawings, CAD models, instructions, etc. The fundamental reason for the emergence of EDMSs in industry has been in the co-ordination of this voluminous amount of data. Whatever the database or hardware solution, the EDMS provides the project organization with a logically centralized interface to the information with adequate retrieval mechanisms.408 A.-P. Hameri and J . N ikkola
3.3.Execution
M anufacturingis often a parallel process to the design and engineering processes. This is due to the need for pilot and prototype testing, which are used intensively in projects, e.g. the LHC to test the feasibility of the technological solution. During manufacturing, the design is rst time tested in rea-lity and the coeæcient in error correction is a dec-ade higher than during the design and engineering. EDMS procedures ensure the manufacturability of each part speci cation by taking into account the constraints of the production facilities. In addition, the EDMS serves as a exible medium to retrieve the latest information related to part manufactur-ing, assembly, quality and material instructions.
Quality management is here referred to as part of the execution phase, although in general the EDM procedures focus on better design and engineering quality through all process phases. The EDM pro-cedures support quality management with continu-ous access to the multiple change logs generated by the diverse processes of the project. The alternating methods of quality management may utilize this information to complement their normal sources of information. This enables the quality management not only to remove problems related to operational activities, but also to improve the whole process within which the actions take place.
Assembly, i.e. the process of tting together manufac-tured parts into a complete unit of a machine, needs to incorporate numerous parts and assembly infor-mation details to perform its task. An EDMS sup-ports this by collecting the required information according to the hierarchical product breakdown structure. Con guration management assures the compatibility of this information.
Installation process is the nal set-up of the system after which the operational phase of the system may commence. To achieve this, multiple information is needed concerning the sub-assemblies and nal assembly site. By taking into account the site and other installation-related constraints, the nal assembly or installation breakdown structure forms another view of the system being built. An EDMS maintains this information in such a form that the con guration management maintains the product coherence with the installation information. 3.4.Exploitation
Operation, i.e. the practical work with the installed system, bene ts from the centralized source oftech-nical information when interpreting the results pro-duced by the system. Well-established technical documentation speeds up the system start-up and provides the operators with an information source for profound understanding of the system.
M aintenance of the system requires well-documented system history to succeed in its task. In a project with long duration and involvement of multiple resources, the detailed documentation of the sys-tem’s change history prepares the ground for smooth and eæcient maintenance operations. The EDM procedures produce a full evolutionary per-spective to the system design and reveal how the system was technologically completed, which greatly helps the maintenance and problem repara-tion work. Also, machine development and physics analysis requires access to the past, current and future view of the accelerator.
Learning, not only after but during the project, is important and an apparent feature of successful long-term projects. EDM procedures produce struc-tured information that enables the induction of exo-genous learning inside the project organization. Reported instructions on changes and tackled prob-lems increase the knowledge and skills of the organ-ization, which eventually helps to avoid making the same mistakes.
System demolition, a major project in itself, will inevi-tably take place at some point. To manage the task in an environment, time and cost-eæcient way, the documented technical details of the system are needed. Thus, the EDM procedures serve the whole life-cycle of the system.4. Experiences with the prototype
At CERN, more than 40 countries are collaborating in a project producing the most complicated scienti c instrumentation ever built by mankind. Engineering and design work takes place on a global scale, resulting in drawings and documents with other project-related information which are distributed through the Internet and World Wide Web. A special WWW-based document management system has been developed to interface between the global community and document databases. The system, called TuoviWDM, has gained momentum at CERN and is being tested in parallel by several com-panies producing one-of-a-kind or make-to-order prod-ucts requiring tailored engineering and design work. The experience shows that WWW overcomes many obstacles retarding the development towards document-integrated collaboration among geographically distributed organi-zations and industrial networks.
One-of-a-kind and make-to-order companies have been concentrating on their core businesses, which even-tually has increased the number of partners in a product delivery process. In many cases, third parties deliver well over 75% of the end product’s value. This kind of distrib-uted and collaborative work requires universal document accesswith low cost and compatibility with diå erent infor-mation technology ( IT) infrastructures. The Internet and World Wide Web forms the universal protocol needed to tackle distributed process across geographical and technological distances.
The TuoviWDM supports distributed design work by providing engineers with a simple and secure means for storing their documents and making them universally accessible to remote collaborators. The system allows one to store, search and retrieve documents through a WWW browser. It is thus suitable to a wide range of document management needs, where user authentication and authorization are needed. During the project ow, an automatic noti cation mechanism informs the appro-priate users about changes, and the project engineer is able to call for progress follow-up reports based on docu-ment changes and accesses. Being a platform-indepen-dent system, the TuoviWDM has been successfully integrated to several in-house and commercial engineer-ing and product data management systems ( EDMS/ PDMS) .
Accessing documents is ne for most of the users, but to understand how the whole creative process is taking place is much more diæcult. To move in this direction, the system supports document control, which enables one to view communication, i.e. accesses, around the docu-ments. This provides both the project management and each individual contributor assistance in establishing a coherent view on how the actual work takes place. There are various ways of viewing the document commu-nication. One may view communication around one single document, e.g. around a document him/herself has produced to check that the contribution has been received by those who are mostly concerned. Also, one may send targeted reminders to those who may have not noticed the document or revision.
On a group level, one may analyse the communication between people. There is an assumption that people accessing the same document share some interest in the same subjects and concerns in the project. This way, a group-communication grid can be formed, where the group may be viewed as a bunch of individuals or domains or even nations. This way, management may view whether communication takes place and how it changes in time. Clearly, lack of communication in a sub-project is a sign of problems in collaboration. By viewing the document control analyses from a time per-spective ( Hameri and Nihtila¨ 1997) , one may reveal
cer-410 A.-P. Hameri and J . N ikkola
Figure 2. A snapshot of the TuoviWDM system in use.
tain milestone-related behaviour, but also that develop-ment work takes place in communication bursts, i.e. dur-ing periods of intense interactions followed by calm periods. Figure 3 shows a snapshot of the system with the document control graph.
The communication graphs take notice of performed changes, i.e. on whether the document has been read, changed and resubmitted. This way, the dynamic fea-tures of the collaborative work can be reviewed. The most used function in the pilot project concerns the con-trol of the vital documents in the projects. These docu-ments are related to project schedule, design parameters and other project-related constraints. This has proved to be very useful to check that each and every engineer is using the same parameter versions on their current devel-opment work. The system supports this kind of operation by, after a given time period, sending automatic remin-ders to those who have not retrieved the latest parameter lists.
It was thought in the beginning that the document control features were understood as means to control the activities of an individual, i.e. a big-brother watching syndrome was expected ( see Hameri and Nihtila¨ 1998) . Fortunately, this was not case, as the tool was welcomed as an additional means to help daily work and to keep the development work on the right track. Again, the intuitive and easily accessible tool, i.e. WWW and Internet, low-ered the threshold to commence the control activity, yet
one must add that there is still a lot to be done in this direction, as the approach is obviously new to many people throughout the project organization. With the above-mentioned key documents, the situation is some-what diå erent, as the function to secure that the message has been received is frequently used.
The TuoviWDM toolset is used by a number of pilot projects. Some of the pilot projects are open for global access*. The system serves several hundred users around the world and it hosts around 10 gigabytes of engineering information. Two in-depth industrial pilots with com-panies performing global design and engineering activ-ities have demonstrated that:
the amount of key documents concerning all part-ners in a product delivery process is surprisingly low, and that better control of these documents already improves the overall project performance;
inside the companies, around 70% of the documents are disseminated and retrieved electronically, yet between third parties only 25% , the rest being delivered by conventional and manual means;
Internet-based protocols together with WWW interface overcome problems related to diå erent* Readers are encouraged to see the system in full use as a guest user from the following URL: http://tuovi.cern.ch/
Figure 3. A snapshot of the document control front-end. On the right is a spider graph on the communication within one sub-project at CERN.
computer platforms, however, focus on document formats is of great importance;
WWW-based TuoviWDM interface has been suc-cessfully integrated into several in-house and com-mercial engineering document systems, thus the technology is exible and highly applicable already;
security aspects with strict user authentication and authorization indicate that, at least so far, the applied medium is highly secure and reliable, even for distributing nancial information.5. Conclusions
Engineering data management supports all phases of the project, right from the conceptualization to the exploitation of the project results. Its importance at the beginning of the project is easily neglected, although it is exactly at that point when the necessary procedures must be implemented and used in order to have all the support for the latter phases of the project. Implementing docu-ment access is not an easy task if it is imposed wrongly by the management. The best way to adapt to the new working method is to start small and think big. The opposite way is doomed, as the users may not be willing to use unfamiliar methods. Experience indicates that the best results grow from need and from satisfying it. This is achieved when the use of the system grows along the design process. The earlier in the project a common net-work net-workspace approach is adopted the more used and bene cial it will be. There are certain managerial activ-ities that have to be agreed before using the system ( Globerson 1994, Bachy and Hameri 1997) . These are related to the structuring of the whole project. The fol-lowing steps may be followed:
( 1) De ne and freeze the top-level product breakdown structure ( PBS) of the system which is to be accomplished by the project. This structure, which equals the bill-of-material in production management, sets the skeleton for further structur-ing and organization of the project.
( 2) Agree on the general classes of the documents that are to be produced, stored and distributed. Too detailed structure leads to misuse and confusion. Existing classes are usually enough to start with, which also makes the initiation to new procedures and systems easier.
( 3) Establish common naming and identi cation con-vention for all project-related documents, but do not make it a burden. Experience shows that as projects are on their way changes are inevitable, too rigid naming conventions that are not
adapt-able into new situations produce extra mess in the document management.
( 4) De ne document status’s that best describe the life-cycle of data, e.g. the development process around the documents. Diå erent document classes may have speci c life-cycle status’s through which they enter during their evolution.
At CERN, the PBS is used as the front-end for brows-ing in the document workspace. Nambrows-ing schema provides the basic structure for nding and identifying individual documents, which also sets the means for controlling vari-ous document versions along the project duration. Status indicators add one more attribute to understand how the project is evolving and which documents are valid, under development or have been changed. In order to fully exploit a project workspace, the following additional de -nitions make the procedure and the supporting system even more bene cial for the project:
( 5) Allocate responsible resources to the PBS, and tie the available manpower under the respective sub-system, component or device level. This sets means to establish virtual project structure available to everybody through the information network. ( 6) Generate mailing lists for each level of the PBS to
enable automatic messaging and noti cation mechanisms to keep the organization up-to-date on all relevant changes to project documentation. ( 7) Associate key design parameters of each and every component, sub-system or system to the PBS. Make sure that all concerned retrieve the tion and that documents related to this informa-tion comply with it.
( 8) In addition to baselining the main project, provide individual workspaces with more relaxed pro-cedures to pave the way for a habit of structured document and data management in the project. These actions pave the way for eæcient communica-tion and resource co-ordinacommunica-tion in distributed project organization. In practice, a project workspace manifest-ing the overall structure of the project helps on an indi-vidual level to see ones position in the overall project structure. Along with the naming convention, linking parameters to the workspace is an essential step towards version management and automatic checking of the system con guration. Although full-scale work ow and con guration management are operations which are to be carried out by dedicated systems, the experience indicates that in distributed projects the WWW acts well as a catalyst towards better and improved working practices.
412 A.-P. Hameri and J . N ikkola
References
Bachy, G., andHameri, A.-P.,1997, What to be implemented at the early stage of a large-scale project.International J ournal of Project M anagement,15, 211–218.
CIMdata Inc., 1994, PDM Buyers Guide ( USA: CIMdata Incorporation) .
CIMdata Inc.,1997, PDM Europe ’97, Conference Proceedings, 28–30 October 1997 ( USA: CI Mdata Incorporation) . Frederick, M.,1995, Case Study: Achieving buy-in. In12th
International Conference on Enterprise PDM : T he N ext Generation, Boston, Massachusetts, 15–16 May.
Globerson, S.,1994, Impact of various work-breakdown struc-tures on project conceptualization. International J ournal of Project M anagement,12, 165–171.
Hameri, A.-P.,1995, Con guration management in project driven manufacturing—guidelines to better performance.
International J ournal of M anufacturing System Design,1, 343–349. Hameri, A.-P., andNihtilÄ, J.,1997, Distributed new product development project based on Internet and World-Wide
Web: A case study. T he J ournal of Product Innovation M anagement,14, 77–87.
Hameri, A.-P., andNihtilÄ, J.,1998, Visualising the factory through meta-manufacturing processes. Production Planning and Control,9, 28–35.
Hargrave, R.,1995, Building a PDM system for immediate value and long range potential. In12th International Conference on Enterprise PDM : T he N ext Generation, Boston, Massachusetts, 15–16 May.
McCrea, S., andD’Agostino, P.,1995, After the pilot. In12th International Conference on Enterprise PDM : T he N ext Generation, Boston, Massachusetts, 15–16 May.
NihtilÄ, J., 1996, Integration mechanisms in new product development. Dissertation, Helsinki University of Technology, Espoo.
Puttre, M., 1992, Document management: establishing an electronic archive.M echanical Engineering, January, 74–78. Sulonen, R., 1995, Product data management.
Hewlett-Packard seminar, Espoo, Finland, June.
