Active Knowledge Modeling

Large projects are complex, involving multiple products, disciplines, methodologies, techniques, systems and stakeholders. They often fail to meet expectations, schedules and budgets, and results are often poorly validated and managed. Current IT support to project definition, planning, execution and management is fragmented and rigid. Lifecycle data exchange is transformative rather than evolutionary. Innovation is not driven by evolution, embedding experiences and lessons-learned. There are many uncertainties and unknown dependencies in the early phases, many nonproductive meetings, stove-piped and sequential information flows, poor data management, and limited knowledge sharing. Collaborative design, cross-functional team working, and service composition are inhibited. Work environments and user interfaces are rigid and discipline-specific.

This post raises the questions: Are we wrongly trying to generalize and program creative work, collaborative and adaptive environments, and human behavior? Are we doing right to standardize properties, embedding their parameters and values in code? Can this approach serve complex customers with dynamic demands for controlling dependencies, supporting innovation, and automating adaptation?

To deal with the complexity of collaboration, coordination and control, projects must be designed. The purpose of project design is to support stakeholder involvement, collaboration, continuous innovation, concurrent engineering, customized delivery and operations. This implies integrating project initiation, planning, design, execution, validation and management.

Most projects innovate and design products of some kind, so design methodologies are already part of the knowledge base of most projects. However, workplaces, services and methodologies for capturing, integrating, reusing and managing project, product and process knowledge are missing.

What is Project Design?

Project design is about capturing, relating, and managing dependencies among business, technology, people and roles, products and plants, production, life-cycle processes, infrastructures, and client services. Performance, customer and operation parameters must be continuously calculated, collected, monitored, balanced and modified to react to changes and deviations.

Project design can briefly be defined as: “Creating a collaborative environment with clearly defined roles and responsibilities for managing objectives, resources, contributions, values, and returns.”

Project design requires a holistic approach pushing stakeholder involvement, roles and responsibilities, and cohesive evolution of results. Powerful viewing and knowledge management is necessary.

Current Approaches

Current approaches are dominated by the IT applications and some twenty nonintegrated project methodologies. The IT application systems are too static and do not allow designers, engineers, business people and managers to capture, control and manage their data. Local methods, parameter values, contexts and experiences are lost. This rigidity and lack of support for pragmatic local competence is most apparent in PLM solutions. The rigidity of IT application systems has led to new IT systems developed solely to remedy the shortcomings of the core engineering application systems. Change and portfolio management are examples of such information management applications.

Most services and data representations in current IT applications are formal and standardized. This means that local nuances, method and parameter dependencies, practical experiences, operational skills, advanced design rules, contextual behavior and innovative ideas are lost. At best these core competitive knowledge aspects remain with the workforce. Services for project teams to capture and reuse this lost practical knowledge are demanding more agile solutions and a more holistic approaches to design.

Challenges

Industry projects are becoming increasingly complex and dynamic. Customers comprise natural resource owners, public and private agencies, developers, operators and a variety of clients and citizens, so requirements change as projects and operations proceed. As a consequence work processes, IT systems, and management practices are falling short of growing demands, such as:

• Planning and coordination is difficult, as there is no common view of progress reports and revised plans. Teams keep local progress records and plans.
• Stakeholder involvement and collaboration is not supported, inhibiting design, concurrent engineering and joint ventures.
• Objectives and goals are defined by management at initiation time, they are seldom revisited, and technical objectives are rarely derived.
• Design and engineering projects are mostly discipline driven, as sharing of business and operational data across disciplines is poorly supported.
• Project stages and gateways are used to manage cost and quality, but also to manage information flow. Responsibilities for data and knowledge sharing are not defined.
• It is difficult to manage dependencies across functions, systems, disciplines, departments, and companies as there is no federated representation of the project aspects involved.
• Smart products and plants must be equipped with services facilitating life-cycle event data collection, and communication to other stakeholders.
• IT emphasizes general procedures, formal methods, and standardized data, at the expense of the innovative design and collaborative engineering processes that foster competitive advantage.
• Adapting even simple aspects of software applications takes significant roundtrip time and the costs quickly outweigh the benefits.
• Data collection, aggregation, presentation and manipulation cannot be adapted to user needs, making data management a manual burden.

Project management, execution and control methodologies must be integrated, adapted and extended with methodologies for product design and modularization, process modeling, task execution, resource allocation, quality management, visual solutions development etc. An enhanced project design approach should automatically create contexts for knowledge management, activation and reuse, for continuously improving and adapting methods, and configuring work processes and user interfaces. Services are needed that enable evolutionary modifications, enhancing methods with experiences and issues, and managing roles and responsibilities.

In the figure below, the triangles contain requirements for integrated operations project design. The dynamic nature of the content is illustrated by the arrows. Current IT solutions are mostly concerned with horizontal data integration.  

Challenges in project design

Planning, control and coordination must become more effective, transparent, and real-time. Planning starts top-down, but as creative work progresses it must also function bottom-up and middle-out, based on validated experiences, real-time deviations and changes, and predictive performance indicators.

Opportunities and Requirements

Recent advances in knowledge modeling and model-based systems engineering enable a knowledge architecture to be built as the project proceeds, and product, plant and production evolve. Future project platforms should provide services that implement:

1. Project knowledge architectures relating external and internal object properties and capabilities
2. Front end loading, through reusable field architectures with best practice project knowledge
3. Integrated operations, through model-driven role-specific dashboards and workplaces, automatically providing context for interpreting large volumes of operational data
4. Predictability, tendencies and proactive change management, through automatic architecture-driven aggregation, analysis and monitoring
5. Services for automatically configuring and managing collaboration arenas with common and role-specific views
6. Lifecycle knowledge sharing and reuse through methods and tools that enable designers to create visual languages for the fuzzy front end, and remove barriers between stages and disciplines
7. Global engineering by creating and adapting networked solutions for joint ventures with zero programming, and extending project teams with experts

Project design must be supported by an agile approach, integrated methodologies and an ICT platform that incrementally captures and aligns customer needs and potential solutions. Project participants must take the best out of their practical work environments, mental models and computer-based models. This requires visual work environments that enable proactive collaboration and behavior.

The figure below illustrates the importance of an agile knowledge architecture in meeting these demands. Investments in the early phases pay dividends that are returned many fold downstream. These challenges and opportunities are similar across industrial sectors and many public services. It is often said that ”What you don’t design for you’ll never get”.

Active knowledge architectures support project design

 
The Active Knowledge Architecture Approach

The challenges and the solutions on demand can be implemented and deployed only if roles and responsibilities for methods, services, tasks and data are clearly defined. Project platforms must support continuous improvement and participative learning-by-doing, transforming pragmatic knowledge and experience into role-specific competences and skills. Visual, model-driven user interfaces will close the gaps between design, execution and business management. Reuse of knowledge architectures is facilitated by filling, repeating and replicating roles.

An Active Knowledge Architecture (AKA) is developed, used, and managed by project personnel applying a model-configured platform. With an AKA, the model is the application. Models define user roles, and behavior and layout of the user interfaces. Any model is pure project data, so they are easily modified by users. Solutions become self-modifying and self-managing. By using knowledge architectures to design projects, stakeholders can share ideas, concepts, sketches and views. Requirements are related to functions and properties. Functions and properties are structured and connected to solutions, processes, resources and methods. Through visual modeling the activities are refined and related to roles, tasks, information structures and operational views.

Project teams can reduce meetings, rework, conflicts, changes, and errors. Misunderstandings caused by lack of overview and insight, are avoided by proactive learning and collaboration. The quality of results is improved, and solution concepts become more precise. The knowledge architecture can be set to automatically capture and analyze data, and configure views for design, engineering, procurement, and fabrication, or any other lifecycle activity. Concurrent engineering and integrated operations have to be designed for dependent on product design methods and rules. Reuse and alignment of solutions across projects are directly supported.

Active knowledge architectures connect, refine and manage different views on key project dimensions:

• Products, systems and components to be designed, fabricated and constructed
• Organizations, project roles and responsibilities, and similarly through operations
• Processes, concrete work plans and processes for quality assurance
• Systems & infrastructures, including fabrication, construction as well as IT systems

Knowledge dimensions for project design

Project design through Active Knowledge Architectures (AKA) brings control to data and knowledge sharing across networks of users and IT systems. Role-oriented views facilitate cooperation, learning and reuse, giving everyone the information they need to perform their work. New services for performing operations on data can be added at any time. An AKA supports evolving design by integrating precise data, concepts and ideas, open to interpretation. Agile environments for creative work execution and management can be designed and configured by users.

Services for capturing and managing complex dependencies across the different structures can be composed. Operational methodologies are offered as templates and patterns that make best practices, guidelines and experiences applicable in different projects. This creates a sound, traceable and predictable base for communication, coordination and decision-making.

Project Specific Strategies and Concepts

In the spring of 2007 we performed a pilot study on project specific technology strategies within offshore field development. The goal was to study how an AKA could improve strategic overview, execution and control of engineering work to improve the design basis. A model-configured workplace for Systems Engineering was developed. A snapshot, intended for the process engineer role, is shown below.

Configurable visual workplace for project design

Bottom right is a product view, a P&ID-diagram. The dependencies between the functional process design and field parameters (bottom left) are captured, enabling rapid changes to the design as field parameter values change. The top-view is an emergent process model, a task pattern, which captures design issues as work progresses. Problems, alternative solutions, pros and cons, and decisions are represented. By linking these to the product model, we capture the design rationale behind it. By adding an organizational model of roles and candidates, we capture who is responsible for what. The AKA can automatically configure workplaces for the roles and people involved. By modeling system and infrastructure components, we integrate existing applications, and views of documents and data sources.

Dynamic progress monitoring and execution

In the autumn of 2008 the AKA approach was piloted in an oil and gas field development project. The main objective was to demonstrate how an AKA-driven solution could implement overviews to improve project execution. The pilot scenario was design completion of pipe-runs in a specific topside area. The entry point of the area manager’s user interface is shown below.

Configurable user interface for team and project management

The AKA structures and governs the many dependencies between systems, disciplines, properties and tasks. The example above illustrates how a piping area manager and his team are equipped to manage progress reporting, resource allocation, quality and work management. The upper left view is an overview of status in each area, while the three bottom views give design completion per week, progress towards the end of the current phase, and to the right we find the backlog of layout tasks per week. The underlying data can be viewed by clicking on a diagram element. The last diagram, upper right, is used for coordination with other disciplines, systems and suppliers. Data is automatically extracted from engineering applications. The configuration of the solution is controlled by model data managed in the active knowledge architecture, without any programming.

Project and work management is a continuous process modeled as roles, tasks resources and views. Work planning and control, progress monitoring and reporting is jointly performed, as is quality management. All is based on real-time overviews shared by the people concerned, or customized for individual roles.

We hope to have provided adequate answers to the questions posed at the start of this post, or at least to have identified approaches to get answers. We believe the concepts and principles behind our approach point to a very exiting future for Project Design, and for Systems and Software Engineering.