Knowledge Management in engineering: supporting analysis and design processes in innovative industries Yvonne BARNARD 1, Arnd ROTHE 2 1 EURISCO International,4 Avenue Edouard Belin, 31400 Toulouse, France Tel: +33 5 62 17 38 27, Fax: +33 5 62 17 38 39, Email: barnard@onecert.fr 2 Airbus Deutschland GmbH, 21111 Hamburg, Germany Tel: +49 40 7437-0, Fax: +49 40 7437-4422, Email: arnd.rothe@airbus.com 1. Introduction Abstract: In this paper we describe knowledge management in engineering. In the WISE project we have investigated the needs of engineers, involved in analysis and design processes in large innovative industries. We focus on the reuse of knowledge, on how knowledge can be found, understood, reused in a new context, and stored. A knowledge portal has been developed which gives a single access to all kinds of different knowledge sources and tools. This application provides support in areas such as structuring the work environment, accessing tools and knowledge sources, contextualising knowledge, annotating, tracing knowledge use and collaboration. Engineers need a large amount of reliable knowledge when designing complex and safety critical products. Engineering knowledge in the industries in which these products are developed resides in documents, databases, design and analysis tools and humans. Reusing existing knowledge saves resources and improves product safety and quality. However, reusing knowledge is not a straightforward process [1]: it will often not be possible to use knowledge created in a certain context directly in a new context. By Knowledge Management (KM) processes, companies create and use their institutional and collective knowledge, organising and controlling the operational processes in the knowledge value chain in the most efficient way. KM focuses on facilitating and managing knowledge related activities. KM is the systematic, explicit and deliberate building, renewal and application of knowledge to maximise knowledge-related effectiveness of an enterprise [2]. In this paper we focus especially on the reuse of knowledge, on the question of how knowledge, in the form of what we call knowledge objects, can be found, understood, reused in a new context and stored. This process is visualised in Figure 1. In the WISE (Web-enabled Information Services for Engineering) project, a European research and technological development project (IST-2000-29280), we are working on two objectives: understanding current engineering practices and identifying the areas where support is needed, and development of procedures and tools for facilitating and improving KM processes [3]. The industrial partners involved in this project are Nokia and Airbus. Other partners are Cyberstream, Interface SI, PACE, EURISCO International, Norwegian Computing Centre, Helsinki University of Technology, and Technical University of Berlin. In the following sections, we will elaborate on the needs of engineers regarding KM. Next we will explain the solutions to address these needs as they are developed in the WISE project. Finally these solutions are discussed as well as further work needed to support engineers in their work practices.
Figure 1: Engineering knowledge in action (KO = knowledge object) 2. Needs of engineers for knowledge management The needs of engineers have been investigated from a variety of viewpoints and with complementary methods, such as interviews, brainstorming sessions, use case modelling. In this paper, rather than describing the methods used for gathering user requirements [4], we will focus on the results. Engineering practices were studied in Nokia and Airbus cases: 1. In the mechanical parts design case the design engineer has to produce a CAD model of a mechanical component. Inputs are strengths requirements from the stress department as well as geometrical and manufacturing constraints. 2. In the electronic system design case, a complex process of producing requirements, specifications and design documents is followed. The focus is on issues of cooperation between different contractors, short development cycles and high employee mobility. 3. In the safety analysis case the safety analyst has to verify the safety and reliability of a system component. For this he/she needs to analyse all possible causes for failure. For each failure cause he/she needs to elaborate and document reliable back-up solutions. In all three cases engineers work on projects. These projects are dedicated to the development of a new product for example the Airbus A380 aircraft or a Nokia base station. Project duration may extend to several years and their size may be very large (e.g. 1000 engineers). Typically, several projects in an engineering company may run in parallel and an engineer may be involved in different projects at the same time. The industrial cases concern different engineering processes, but finding and reusing knowledge is important in all three of them. Different accents on the needs for support are: Reuse of designs of (parts of) systems (Figure 2). The designers want to find parts that have been already produced in previous projects, which have the same characteristics as the one needed for the current product, or which only have to be slightly modified. If old designs can be reused, this not only saves time (and cost) during the design phase itself, but also in the verification procedures of reliability and quality and in the production process, because already existing manufacturing tools and jigs can be used. However, because they are created with CAD (Computer Aided Design) tools, it is not always easy to search for designs with their specific characteristics with conventional search tools. The context of the previous usage needs to be understood.
Figure 2: Reuse potential in mechanical design Figure 3: Learning from previous design situations Advise on the design process (Figure 3). In the knowledge intensive industries we are dealing with, huge amounts of expertise and knowledge exist, in databases, in documents, in the heads of people, often explicit, but sometimes only tacit [5]. Knowledge is also incorporated in the designs themselves. Engineers have a strong need to get access to these knowledge sources and to be able to make tacit knowledge explicit. So, engineers need support to search for knowledge, to find the right source and to extract the knowledge in a form that is usable for the current task. People work together with colleagues at different locations and with other partners and subcontractors. Therefore knowledge is spread out over and beyond the company, and engineers do not have the time for extensive searches or for acquiring all knowledge themselves. Support is needed for to be able to make use of the experiences of others and to build upon expertise accumulated in the past. Figure 4: Using previous knowledge in Safety Analysis Validation, verification and justification (Figure 4). In safety and reliability analysis of designs, as well as in the design of safety critical systems itself, engineers have to make sure that they use all the available knowledge, such as lessons learned from the past and best practices established within the company. They have to check the designs against all knowledge about problems, incidents and accidents in the past in order to make sure the design conforms to the norms and to exclude all foreseeable problems. Furthermore, they have to justify the design choices and to explain the design rationale. These verification and validation procedures have output that is used in further processes, also outside the
company, such as certification procedures. Engineers need support that helps them to justify their design in an easy manner, and to make sure all available knowledge has been used. For all three industrial cases, the organisation of the work is an important topic. Because of the complex structures and procedures inherent in design and analysis projects, support is needed for both team leaders and team members in setting up the work, the collaboration procedures, the sharing and protecting of documents, both final and draft versions, finding the knowledge sources available etc. 3. The WISE knowledge portal Engineers in these innovative industries do not need another tool; they are well-equipped with documentation and specialised tools. However, knowledge is fragmented located in different places, and a variety of dedicated tools are used. Two concepts are central in solutions for the management of engineering knowledge: 1. Knowledge objects: engineering knowledge resides in documents, databases, design and analysis tools, humans etc. Whatever the location or the format, engineers are to be provided with this knowledge with a single access mechanism. Knowledge objects can be searched, annotated, traced and shared with others. 2. Reference persons: knowledge and experience from other persons are of high value for engineers. They can come from experts, but also from (distant) colleagues, subcontractors, or anyone on the company s network. Knowledge sharing amongst people is not a matter of the transfer of knowledge from expert to novice, but of sharing between people who work or have worked on relevant processes. The solution we have developed is that of a knowledge portal, not replacing existing tools but offering a single entrance to all sources of knowledge. It also offers support for understanding the complexity of knowledge and for achieving information safety and reliability both important aspects in engineering projects. The WISE knowledge portal was developed as a basic platform, which can be customised according to the specific needs of very different industrial cases. The software is entirely built using industrial standard or de-facto standard technologies: Java, Oracle, JDBC, Oracle application server. Figure 5: WISE system architecture There are three tiers in the WISE architecture (Figure 5): The client layer enables user interaction with the WISE platform directly through their legacy applications or through a
web browser. Therefore, a presentation management module provides services related to supporting several types of clients and delivering the functionality of WISE on different platforms/devices. The knowledge server layer carries out all information processing, using user input from the client interface and persistent information that is retrieved from the system s K-base layer and pushed to the client layer. The knowledge base layer includes the WISE Meta database which stores all data related to the functions of the WISE platform and a middleware layer which enables access to distributed engineering data in legacy repositories. In Figure 6, a screen dump is given of the WISE portal. The first prototype has been finalised and tested within the three industries. At the moment, a second prototype, an improved and extended version of the first prototype, is being developed. Figure 6: Screen dump of WISE knowledge portal, annotating knowledge objects 4 Support for engineers in knowledge management The main areas in which KM in engineering is supported by WISE are: structuring the work environment, seamless access, contextualisation of knowledge, annotation and validation, tracing knowledge use, subscription to knowledge objects, and collaboration. 4.1. Structuring the work environment Engineers work in very complex environments and projects, with a variety of tools and knowledge objects. We support them with a single portal to all these sources, and a workspace and workflow that can be contextualised and personalised, by the engineers themselves and /or by team leaders, both on an individual level and on a project level. The engineer is presented with information that is relevant to her/his current work context, e.g. the product being developed, his/her role in the design process, the current phase in the development process etc. The team leader can create the context for the project, such as persons working on the project, their tasks, roles, workflows, deadlines, access rights etc.
The team can define which knowledge objects, tools and other knowledge sources are relevant to the task and put them as links into the workspace. By this contextualisation process, the team ensures that all knowledge objects the engineers must use are indeed available. With this functionality we address the need for support in the organisation of the work, as expressed in all industrial cases. 4.2. Seamless access to all relevant tools and knowledge sources The single WISE portal gives access to the dedicated tools, such as CAD tools, and to all databases accessible to the engineer. The tools can be launched from the workspace and data can be imported and exported. The WISE portal guarantees interoperability and provides a rich set of search functionalities. WISE does not store the knowledge objects and data themselves, but only metadata and links. Thus the integrity of the companies informational infrastructure is guaranteed and the function of WISE as a portal is realised. WISE is able to search in all kinds of knowledge objects. On a query from the user, WISE suggests a list of possibly relevant knowledge objects, together with some information, as it is contained in the meta-data such as source, type of knowledge object, summary, etc. The user can select a knowledge object and open it. This single access facility makes it easy to find, and thus to reuse, knowledge available within the company. 4.3. Contextualisation of knowledge objects Knowledge objects, such as designs of system (parts), are tagged with XML meta-data, thus becoming searchable. Knowledge objects produced in a certain context, for example in the detailed mechanical design phase, will be automatically tagged with this context and will thus be made available to other engineers in this phase, in the same or in another project. Templates can be created and customised for different types of knowledge objects, with pre-defined attributes. WISE can either assign metadata automatically, such as the context and the name of the person who modifies the object, or other metadata can be added by the user. This feature supports the user in understanding the context in which a knowledge object was created and used, thus facilitating reuse. 4.4. Annotation and validation Annotation facilities for knowledge objects are provided, so that engineers can immediately write comments and justifications while working on a task. This is important for three reasons: to capture the design rationale at the moment that it is produced, to prompt engineers to explain and justify their choices and to enable the communication between engineers and with other stakeholders. The validation and verification processes particularly profit from these facilities; it makes it easy to check knowledge objects created or modified by engineers and to communicate if things are not clear. By annotating, knowledge objects become active sources of knowledge, not just static pieces of information. Engineers add their experience and knowledge, so they may benefit others. Annotation facilitates the cooperation between team members, but also makes it easier in the future to understand why certain design choices or recommendations were made. In Figure 6 a screen dump is given of the annotation facility. 4.5. Tracing the use of knowledge Traceability of design and safety analysis processes is crucial in the highly advanced and safety critical industries we are addressing. We developed a functionality that builds and visualises traces on the use of knowledge objects. The setting and use of traces can be
customised on several levels. This tracing capability is especially important for the two focus areas described in Section 2 on verification and advise. The traces make it possible to automatically determine whether the engineer has indeed consulted all the obligatory knowledge objects, and for which purpose. For example, to analyse the safety of a certain system, the engineer should have reviewed all the incidents related to that system. Tracing not only enables checks of completeness, but it also helps the engineer to make the justification, by having an automatic link to the source upon which he/she based a recommendation about the safety and reliability. To get advice, the engineer can find out what kind of knowledge use behaviour other people in the same role have shown by following their traces and finding knowledge that may be helpful. The engineer can also contact the person who worked with a certain knowledge object and ask for personal advice. In interviews with engineers, it turned out that there are different feelings about tracing the search and consultation behaviour. For some groups of engineers, tracing is seen as beneficial, because it helps to be transparent. For others, tracing may be a threat, because other people can criticise premature work, and thus hinder creativity and make people (especially inexperienced engineers) afraid to make errors. What to trace and who to give access to is a process that needs careful discussion amongst engineers and their managers. 4.6. Subscription to knowledge objects and active notification of changes In innovative design things change quickly. It is of the utmost importance that engineers are aware of changes and always use the latest version of a knowledge object. Engineers can subscribe to knowledge objects and be notified of change. This notification can be passive (announcement when opening the workspace) or active (by email), and different forms of subscription and levels of notification are available. The user can use several parameters to customise this notification process. For example the user determines whether he/she is notified when the knowledge object is updated, deleted, given another status etc. By making it easy for the engineer to be sure about the status of a knowledge object, knowledge sharing is facilitated. Dealing with draft knowledge objects, such as reports and designs, is a major problem in all three industrial cases. It may be dangerous to use knowledge which has not been officially validated, because it may contain errors and be incomplete. However, sharing early drafts is beneficial because others can supply feedback at an early stage. In projects, time schedules can be very tight and if drafts are available, others can use them as input and start their work without having to wait for final versions. 4.7. Collaboration WISE provides engineers with a transparent view of the work of colleagues and other relevant persons, past and present. The WISE portal provides a view on shared knowledge spaces, but in a personalised manner, allowing the engineer to keep knowledge private, or open only to team members, or even to other groups. Knowledge is often situated, created in collaboration between humans in a specific context. WISE supports this collaboration process by providing facilities for searching reference persons and by a Frequently Asked Questions (FAQ) facility. How people are described in the list of reference persons remains an open question. We have found that legal rules and privacy and company policies make it sometimes difficult to have yellow pages with extended information on a person. For example naming someone as an expert can have far-reaching implications, e.g. for his/her position in the company. Also not everyone is willing to be a reference person and have to answer questions from others. However, finding the right people is an important need for all engineers in our industrial cases. Facilities such as FAQs need to be carefully organised in order to avoid incorrect answers being made available. Again, verification and validation procedures, as well as a definition of responsibilities are needed. In the industrial cases we
address, collaboration is not only a matter of informal contacts and the exchange of ideas, but it may have also serious implications for safety and reliability. 5. Conclusions In this paper we described the needs of engineers in engineering processes, and the solutions the WISE project provides, solutions both in terms of software and knowledge processes. KM for engineering is becoming more and more a necessity in fast-moving industries dealing with extremely complex processes and products, and in a highly competitive market [6]. Engineers who have access to knowledge that helps them to do a better job, are of high value to their companies, not to mention the importance of personal satisfaction. Designers do not want to re-invent the wheel. If they can find knowledge on which to build further, they can provide a higher level of innovation in their product. Both the knowledge gained in the project and the software developed will be exploited. Software development is an important part of our work, first for the development of a prototype product that will evolve into operational software actually used in the industries involved as well as into products for other engineering organisations. Secondly, the development and testing of the software also brought us more insight into the KM processes in engineering. This knowledge will be used in the industrial cases to improve their KM processes, and to improve and extend the KM applications already in place. Knowledge is not the same as static information, stored in files and databases. Knowledge has to be kept active, or alive, by applying it in specific situations, by discussions with other people, by adding new experiences etc. The point of view of the WISE project on KM is to facilitate this keeping knowledge alive process. An important step in this process is to make knowledge and its use transparent to other people who can reuse and build upon it. In Figure 1 we depicted the situation of an engineer who reuses a knowledge object from one context to another. For KM to be effective, this situation should be linked to the situation of other people, other tasks, other knowledge objects etc. etc. KM should be seen as a process over time, storing knowledge for future use. To spread out KM over such a network is the challenge of innovative industries and the topic of our further research. How to reach this aim is only partly a matter of adequate software, but far more a matter of organisation [7] and consensus between the people involved. References [1] Blackler, F. (1995). Knowledge, Knowledge Work and Organizations: An Overview and Interpretation. Organization Studies, 16/6. [2] Kemp, J., Pudiatz, M., Perez, P., & Munoz Ortega, A. (2001). KM Technologies and Tools. D1.1 KM Framework. European KM Forum (IST-2000-26393). www.knowledgeboard.com. [3] WISE website: www.ist-wise.org. [4] Barnard, Y., & Blok, I. (2003). Gathering user needs for knowledge management applications for engineers in advanced manufacturing industries. In S. Bagnara, A. Rizzo, S. Pozzi, F. Rizzo, & L. Save, Proceedings of the 8 th International Conference on Human Aspects of Advanced Manufacturing: Agility & Hybrid Automation, p. 339-346. Rome: National Research Council of Italy, Institute of Cognitive Sciences and Technologies. [5] Nonaka, I., & Takeuchi, H. (1995). The knowledge creating company: how Japanese companies create the dynamics of innovation. Oxford, UK: University Press. [6] Johnson, M. (Ed.) (2002). State of the art survey: managing engineering knowledge. WISE deliverable 1.1, Helsinki University of Technology. [7] Gao, F., Li, M., & Nakamori, Y. (2002). Systems thinking on knowledge and it s management: systems methodology for knowledge management. Journal of Knowledge Management, Vo. 1 No 1, pp. 6-14.