Feature

posted 27 Jan 2004 in Volume 7 Issue 5

Pushing KM into top gear

Managing knowledge has been a priority at Renault since knowledge systems were first designed in the early 1980s. Jean-Marc David plots the evolution of KM projects at the automotive company and describes the how the foundations have been laid to roll out KM at the enterprise level.

Like most industrial companies, managing corporate knowledge has been an ongoing concern for Renault in one way or another. In the 1980s, knowledge systems were designed to capture our expertise in, for example, design or diagnosis (for vehicle or manufacturing tools). Practitioner networks have also been deployed to allow specialists across different plants to share problems and good practice.

In the 1990s, Merex, a method for organising our return of experience or lessons learnt, was designed and applied at Renault, and used most notably in R&D. Domains of strategic expertise were identified, together with the nomination of Renault experts, which led to our first map of critical knowledge.

More recently, the Knowledge Management department was created to deliver KM projects and provide their infrastructure and principles. Renault has identified KM as a domain of strategic expertise.

This article describes the development of our KM projects and examines the issues we faced in deploying KM within a large company.

Know-how management

Renault’s organisational structure distinguishes between domains (technical domains and domains of expertise, for example) and projects. Technical domains (such as brake-system design and crash validation) are responsible for bringing together the knowledge necessary to execute a project (see figure 1). This structure is used within our R&D department, which develops vehicle or engine projects, our IS/IT department and many other areas.

The objectives of such a structure are to shorten development lead-times by avoiding trial and error in the project-development process. It also ensures projects apply best-in-class solutions, learn from experience to avoid the mistakes encountered in previous projects, and standardise solutions from one project to the next to reduce diversity.

Know-how-management projects are based on the recognition that the documents and content that support the organisation are, to varying degrees, incomplete, in some cases still paper-based and are inconsistently saved, shared or up-dated. As a result, users (typically project teams) are unable to access knowledge. They don’t know if the knowledge they seek exists, or if it should, and if they do find what they’re looking for it is unknown whether it is the latest or validated version.

The objectives for know-how management at Renault are two-fold: to enable technical domains to better manage their know-how and to give users better access to knowledge. Know-how management is not just a question of implementing tools, such as a document repository or search engine, it is a holistic system that has to be based on actors, processes, motivation and, finally, tools.

The R&D Knowledge Base

The R&D Knowledge Base is a typical example of know-how management. The system aims to improve knowledge management for the approximately 300 technical domains in the R&D department and to ease access to this knowledge for project teams. Once the roll-out is complete, more than 10,000 users will be able to access the R&D Knowledge Base to improve our car designs.
The project started in 2000 with an exploration phase, which formulated and defined the problems to be solved, identified champions to work with, built a vision and made it concrete with mock-ups.
As a result, the project was officially launched and staffed by the Vehicle Engineering division at the start of 2001. The project’s motto was ‘act first, react quickly’:

Act first – rough prototypes were rapidly deployed to channel local initiatives and were the only way to identify and record users’ needs;
React quickly – Closed-loop iterations of prototypes brought visible progress to the users.

The R&D Knowledge Base has been in use since 2002. It is officially part of the R&D project-management methodology and has also been adopted by other local solutions, first to the Mechanical Engineering division and then the Light Truck Engineering division.

Assessing value

The driver for this project was to enhance engineering design. The strength of our designs has a direct impact on late modifications as changes made once manufacturing tools have been produced can be very costly. The goal of the project was to drastically reduce these costs through right-first-time practices.

By analysing the problems encountered in previous vehicle projects we were able to estimate the ratio of errors caused by failing to deliver know-how to projects or applying it incorrectly. This study has enabled us to demonstrate the project’s return on investment, as well as set up and commit to measurable objectives that will be used to assess the actual value of the project.

Organisation

One of the project’s first goals was to identify the technical domains, nominate and confirm technical leaders. The role of technical leaders was also reaffirmed: “technical leaders are responsible for the R&D performance in their domains. Technical leaders develop innovation to satisfy customer expectations, capitalise on return of experience from projects and are responsible for technical policy.” (Excerpt from the R&D technical-leader mission)

Processes and motivation

Processes and enthusiasm or motivation are essential for ensuring knowledge is up-dated continuously by the domains and applied effectively in projects. The involvement of technical domains in projects has been carefully defined to ensure knowledge is applied correctly. In particular, technical leaders are responsible for the validation of project reviews (for example, they must check that design rules have been applied correctly).

Technical leaders must support projects by applying knowledge, collecting lessons learnt and facilitating tacit exchange between projects by, for example, organising meetings between two projects.

Management’s role within the technical domains was to establish the priorities and ensure that they were met effectively. Priorities, such as determining what types of knowledge must be managed and which are of less importance, are important for keeping KM on track and the system aligned with business objectives.

A support team was also created for the network of technical leaders to train them in how to prioritise rules, for example, and to support them generally in applying these processes. This team can also provide help in document writing or document-quality control (content and indexation) if necessary.

Tools

A document-management application was developed to organise and manage R&D knowledge, and provide easy access to content based on user profiles. Most of the work was again focused on the organisational side to define document types common to all R&D domains, build taxonomies, define user profiles, etc. Management tools have also been developed to monitor the quality of the knowledge base and its use. For example, technical leaders are now able to check their documents and look for obsolete versions.

Lessons learnt

The approach described above is being used to manage know-how in other areas at Renault. It is therefore especially useful to capitalise on the lessons learnt from this project and similar ones. Most of these lessons are not new to KM experts, but they are the ones that are discussed with management prior to the launch of a project or with the team during the project.

Project prerequisites

The first prerequisite is to ensure that management has stated precise project objectives with clear priorities and boundaries. A common misunderstanding is that to do KM you simply accumulate every piece of knowledge in the hope that it might one day be useful. On the contrary, KM must be considered an investment: why should you spend time and money collecting, organising and up-dating knowledge? What types of knowledge should be managed? Know-how management projects that do not specify limits are destined to fail.

Management must also support the project and make it clear to employees that KM is part of their personal and individual objectives. Strong management support will ensure that if difficulties arise, such as change management, there is no reason why they cannot be overcome.

During the project: change management

Three main types of population can usually be identified within these projects:

‘Pioneers’ have ideas and visions, and are willing to bring about change. They should be associated with the project from the start since they have a lot to offer, otherwise they might explore their own ways of managing knowledge;
‘Followers’ progressively accept change, mostly because they can be convinced by rationale argument or by management;
‘Reluctants’ often have many reasons not to accept change. They have to be carefully identified and managed accordingly.
‘Act first, react quickly’ has been a good strategy for creating momentum, especially when goals are still unclear and multiple initiatives have to be channelled.

The progress loop

One of the most difficult questions we had to address was to ensure the knowledge-management system continued to run smoothly once the initial enthusiasm had died down. The operational shift from the project team that had designed the system, to the employees that would run it full time was especially challenging. The approach we have adopted here is to continuously monitor the quality of the knowledge base and its use, and measure user satisfaction and value to align the system with business objectives.

Replication of good practice

Know-how management is just one of Renault’s KM projects. Another important area is the replication of ‘best’ practice (we prefer the term ‘replication’ to ‘exchange’ since the end goal is to re-use good practices).

We have a generic process for replicating good practice, which involves a series of steps: detect and collect, qualify and validate, capitalise and replicate, and measure. For example, the ‘detect and collect’ step involves:

Ensuring that individuals recognise when they have done something worthwhile that could be useful to others;
Systematic analysis of best-in-class plants or subsidiaries, for example, to collect best practices;
Problem-driven search that allows staff to find the person that can help them;

Capitalise and replicate:

Capitalising on re-usable practices;
Pushing practices that have been qualified as ‘very best’ practices.

We have three approaches for replicating good practice. In each case, actors and roles, processes and tools must be defined.

Top-down approach

In this case, the project is decided, promoted and monitored by senior management. For instance, Renault’s Commercial division for Europe has decided to organise the
re-use of best practices between countries.

Horizontal approach (networks and CoPs)

Renault has a long history of promoting networks where peers can share problems and knowledge, especially in the manufacturing area. We are currently working to support communities of practice by providing guidelines on, for example, how to launch a community and encourage participation, and use collaboration tools as members are often spread between sites and across countries.

Bottom-up approaches: building on employee initiatives

For over ten years, Renault has actively recognised and encouraged employee initiatives and creativity. Any employee can make suggestions to enhance his/her own activities or solve a problem. All levels of management are involved in the promotion and recognition of the process. One example is our annual conference, chaired by Renault’s president, Louis Schweizer, where we showcase the best initiatives of the year. Demonstrating initiative is an objective for each employee; incentives are offered and based on the number of suggestions that have been applied and the related savings generated. As a result, 350,000 suggestions are collected each year, and in 2002 Renault saved c57m globally (this is a conservative estimate due to the way savings are measured).

The Employee Ideas Online project builds on and enhances existing initiatives. One goal is to make the process for submitting and analysing suggestions more efficient and transparent. The second is to improve how suggestions are shared across the enterprise so that the best ones can be easily copied and replicated (pull mode), while also detected and promoted (push mode).

We have been testing a prototype for two years and will start rolling it out across the enterprise in the first quarter of 2004.

Other KM projects

Other KM projects are at different stages of development and deployment. They deal with:

Capturing knowledge: using methods such as storytelling we are capturing and sharing the lessons learnt from a change-management project before rolling it out to other departments;
Search, classification and text mining (for business intelligence in particular): analysis of large amounts of documents and information, such as patents; automatic classification of documents;
Knowledge-based innovation methods (such as the theory of inventive problem solving) and associated tools.

Enabling KM projects

The role of the KM department at Renault is two-fold: to deliver KM projects for
specific business needs and to provide the infrastructure and principles for rolling it out.
In the latter role, the KM department has to organise the re-use of solutions from one project to another. In some cases, we are now able to propose generic approaches that are based on various experiences. This is how we developed our initiatives for know-how management and the replication of good practices. We have also built foundations for KM based on three principles: people, document management and collaboration.

People

Renault’s approach to KM has always focused on people. Our goal is not to build repositories or vaults to capture the necessary expertise for a task but to make it easier for people to exchange knowledge, as illustrated above.

Knowledge management here has benefited from other activities, such as the Business to Employees programme, which has delivered an employee portal called Déclic, and addresses change management issues for the adoption of a new culture.

Document management

Two years ago, Renault decided to build an enterprise document-management system (EDMS) to ease access to information, ensure the quality of documents through validation and versioning, avoid duplication of content and documents, and facilitate the re-use of information among processes and publication channels. This EDMS is a key element for the roll-out of knowledge management and the e-transformation of the enterprise.

Documentum’s document-management system was selected as the key element of our technical infrastructure although it offers more than this. Renault created a document-management rule that describes the main principles behind document management in the company. In some cases the rule imposes principles and in others it proposes them to guarantee coherence and avoid diversity. For example, it offers rules on when to create a new document-management application, roles and processes for managing documents, standard attributes, workflow, and so on.

A network of document-management leaders has been established. Document-management leaders are attached to the main department at Renault. Their role is to build a map of document-management applications in their respective departments, identify and prioritise needs, and participate in the governance of the enterprise document-management system, such as evolving the document-management rule.

Finally, change management and communication are also part of the EDMS project to promote better practices and support a document-management culture among Renault employees. Document management is analysed for each major domain, such as product design or product validation, through four main types of documents: reference documents; documents produced by projects; documents produced by micro-organisations; and those from external sources.

Supporting collaboration

Knowledge management requires people to interact and exchange problems and knowledge among networks, communities of practice and problem-solving taskforces. Supporting collaboration across the extended enterprise is a key element for rolling out KM.

One of our objectives is to enable groups, teams and communities to build their own shared workspaces that can be customised, modified and re-used according to their evolving needs and without the need of an IT team. Supporting these groups takes more than just technology; we must also provide methods, guidelines and good practice on how to launch and develop a community.
Previously we used in-house tools for collaboration, however Renault and Nissan recently decided to use eRoom from Documentum and iMeeting from Interwise for improving teamwork and collaboration internally and with our suppliers.

Taking stock

In the past three years, Renault has worked with knowledge management in two ways. Major projects have been launched, such as the R&D Knowledge Base, which have demonstrated the value of managing knowledge to the company. We have also built the foundations for enterprise-wide KM. We are now able to plan the next steps for a more ambitious roll-out of KM that will contribute to the e-transformation of the enterprise.

Jean-Marc David works in the Knowledge Management Department at Renault. He can be contacted at jean-marc.david@renault.com