Realizing Enterprise Knowledge Management
September 16, 2009
Knowledge Management (KM) ranked high on corporate manager agendas in the early 1990s, but KM rapidly became a confusing term that managers scorned. KM systems never delivered what IT providers promised. However, some advanced information management methods were invented, and enterprise portals were developed.
Web 2.0 and Enterprise 2.0 technologies are providing new means for social networking and sharing of personal and public knowledge. However, the core situated and work-centric enterprise knowledge can not yet be expressed, shared and managed by these technologies.
Managing work-centric and situated enterprise knowledge demands fundamental rethinking of not only the nature of enterprise knowledge, but also the practical approach to realizing knowledge management. Knowledge workers must be equipped with services to manage personal as well as role-specific knowledge. Practitioners must be empowered to react to operational events, and become responsible for their own actions, data, workspaces and work plans. Personal competence and skill profiles must match the roles the person is authorized to perform.
This post looks at how knowledge management can be implemented across projects, stages, systems and disciplines, and lifecycles, using active knowledge architectures to update and configure workspaces of work-centric and situated knowledge.
Capturing Work-Centric Knowledge
Workers cultivate knowledge, competence, skills, and experiences as evolving patterns in their mental models. The models are reflected in the way their respective work environments are organized. Anyone who has performed practical work knows of the mutual dependencies among the work process, the available infrastructure and tools, the expected outcome, and the capabilities of the operators. These are vital knowledge dimensions of any workspace, but when working with physical entities we rarely think about the dependencies. However, IT-supported collaborative work need to represent and handle these dimensions explicitly as work-centric knowledge.
Work-centric knowledge emerges in shared environments, when roles produce mutually dependent results. The roles share common views of the work environment and many reflective views of each others contributions. The planner plans, the doers do, the users approve and operate, and managers control, and all would work in concert if they could share context and views. Visual environments or workspaces, built in Active Knowledge Architectures (AKA), are the true enterprise integrators, supporting proactive roles, and shared, reflective views. Work-centric knowledge allows workers to leverage competence and automate knowledge management by executing work in AKA-configured workplaces. Knowledge workers are able to build collaboration arenas, compose and manage services, and transfer best-performance solutions.
Programmed IT application systems can only handle logic and methods in separated knowledge dimensions, and so few of the dependencies mentioned above can be dealt with. Roles can not be defined, and there is no support for capturing context. With AKA-configured workplaces workers are able to externalize, share, improve, transfer and reuse knowledge as competence and skills. Through team working most of the otherwise tacit knowledge can be externalized and captured.
Managing work-centric knowledge implies automatically capturing and sharing data produced when executing tasks. Consequences for other roles can be directly shared and resolved influencing their work in order to respond to the issues raised. This awareness creates agile reflective workspaces, and minimizes change. Proactive knowledge management replaces reactive change management. Software acts as a multi-layered dynamic and reflective medium enabler, building and operating the AKA.
Removing Barriers; Disciplines, Stages and Gateways
Industry is generally organized in distinct disciplines, each with separate approaches, platforms and methodologies for innovation and product development. This rigid segregation of disciplines has resulted in huge interoperability problems, but maybe the biggest loss is the prevention of situated learning and collaboration. Let us explain with an example from industry.
The example is from the oil and gas industry, but construction and public utilities have similar characteristics. In oil and gas field lifecycles key business processes are performed as disjoint projects with limited knowledge sharing. The situation is illustrated in the figure below.
The core business processes above should ideally share most data and knowledge about reservoirs, wells, and the solution concepts in order to get maximum value from the field. Instead core project knowledge remains with the workforce and can only be aired in meetings. Information is captured in proprietary systems and documents, and contents are not easily communicated.
The figure above illustrates another major shortcoming with present approaches. The project views shown are management perspectives. Stages, gateways and milestones are defined’ and maintained for project management to control progress, costs, and resources. All disciplines plan and deliver according to the same milestones. However, collaboration and knowledge sharing among engineering disciplines, product fabrication and other phases are poorly supported. Project plans are not adjusted to account for needs and changes coming from engineering, suppliers, fabrication shops and operators.
There is no IT support for collaboration, collection and sharing of project knowledge across stages, gateways and companies involved. This stove-piped approach makes cost and quality control extremely difficult, and project execution and operations ineffective. Adaptive operations and preventive maintenance will require a holistic approach to project design, execution and management.
Knowledge Management Pilots
In cooperation with industrial companies we have built pilots to demonstrate the capabilities of Active Knowledge Architectures. Pilots validate the ability of the approach to meet customer expectations, secure involvement, raise mutual understanding, and nurture ideas of new ways of working.
The first pilot for product design was developed in the MAPPER project. The customer was Kongsberg Automotive, and the product was their car seat-heating system, as depicted in the upper left view below.
The pilot involved modeling and configuring five different workplaces. The workplace below is called the Configurable Component Builder, and is configured for design of heat producer product components and elements. The upper right view shows modeling of heat producing elements and their parameters. The lower left view supports the modeling of variant rules to automatically configure product component structures based on variant parameters.
The pilot eventually covered most roles involved in the design of the seat-heating system, supporting collaborative knowledge sharing among Kongsberg Automotive and their suppliers. The Kongsberg Automotive pilot demonstrated the capabilities of the AKA approach and methods in replacing supply-chains with collaborative engineering and knowledge sharing. The model-driven configurable components approach to car platform design was successfully implemented. The pilot demonstrator was produced for Swedish Television and their Knowledge Channel.
A second pilot in oil and gas was built in cooperation with Aker Solutions. Project execution dashboards providing overview and insight of work progress are configured to manage collaboration and change. The screenshot below shows how a team manager can monitor status and progress on tasks assigned to his team, and share progress views with his team members and others. The pilot case is piping layout design. The dials show status of tasks assigned relative to the next gateway. The three lower views show production in volume per week, timeline for progress towards completion, and backlog over planned but not yet completed tasks. The upper right view is for coordination among teams. This view summarizes tasks performed by others, which produce results that the team in question is waiting to receive, and thus delaying the team’s progress.
Work plans and progress data are automatically collected from different IT applications, but can also be directly input by the engineers. New workplaces are configured from models in the active knowledge architecture. Updates are performed from the workplaces and thus no programming is required.
Workplaces and views are tailored fit to individual roles, tasks, skills and preferences. Trained users can therefore easily adapt their own work processes, services, and user interfaces.
An active knowledge architecture captures and structures the critical dependencies among disciplines, roles, and tasks. The architecture generates model-configured workplaces and dashboards for monitoring progress. The overall approach in these pilots and operational solutions is depicted in the figure below. IT application systems, the islands of automation, are integrated by an AKA.
The AKA is built by teams composed of industrial designers, discipline and modeling experts applying visual modeling and workplace tools.
Managing Situated Knowledge
These pilots demonstrate a new way of semi-automatically expressing, sharing, maintaining, cultivating and managing work-centric knowledge. Smart links from agile workspaces to more standardized and common database contents were also created. In this way we can turn remotely managed data and knowledge into situated knowledge, linking it to a context that users can manage and exploit.
Project experiences, issues and ideas are added as tasks to emergent work processes. This means that information and local practice is transformed into situated knowledge for easy recovery and analysis. Core knowledge is managed by the owners, that is people performing regular project work. Many wasteful operations are removed, such as entering data in other systems and coordinating solutions among roles. Workspaces providing context for roles are easily replicated, and reused for training and verification.
Implementing work-centric knowledge management and knowledge architectures will allow us to turn all other useful information into situated knowledge, much like using an advanced search engine. The viewing capabilities of an AKA are prepared for this sort of global view content selection.
Summary – Opportunities
Knowledge management of our mental models is automatic. Data is collected by the senses, interpreted by mental models in our memory, and our long term memory is managed by the brain’s motor centre. With AKAs computer models can be used for encoding, collecting, structuring, and viewing, and for recreating situations and contexts that we would otherwise loose conscience of.
Management services for capturing work-centric knowledge and defining responsibilities for project governance, monitoring and control can provide major cost reductions and give more effective projects and solutions. Project management needs to agree on collaboration and management workspaces across the more traditional roles within project and community networks. The organizational impact of role-oriented work-centric knowledge management will be discussed in a forthcoming blog.
Business contracts and other public documents will prevail for a couple of decades more, as will shared community systems such as systems for handling part catalogues and established methodologies. However, the approach to realizing work-centric knowledge management is at the same time a new approach to developing operational IT-enabled solutions, supporting holistic design and enabling life-cycle collaboration and reuse.
The most competitive enterprise knowledge will still be in our mental models, composed of reflective views of customer needs, product and services delivery, and aftermarket opportunities, but we will be supported by explicit work-centric knowledge managed in active knowledge architectures.