Part IV – Discussion and Conclusion
11 Summarizing Research Findings
The findings in paper D confirm that high variability exist in current building design processes. From a design thinking perspective, this could be acceptable if the goal is to increase the understanding of the design problems to solve. As identified in paper E, there is a risk, however, that non-value adding design iterations will occur too frequently if the variability is not carefully managed. Building a strong community within the design team is vital to manage and reduce variability as it allows project managers to entrust the team to find solutions and coordinate activities more efficiently.
Appling the APM methodology fits into this design practice and adds structure to the design development process. Only when such applicable management practices are in place, digitalization can add proper value (cf. Section 1.5). For digitalization to add value, efficient information management also needs to be in place, which requires that information can be captured, structured and exchanged in a standardized way (cf. Paper B and F).
The combined information model described in Section 7.1 includes four information models. In Paper B, the process and product model is identified as relevant and in Paper E, the mission and function model is identified as relevant. PLM systems are already in use in the manufacturing industry and support the integration of the four identified information models. The goal for the AEC industry should clearly be to implement solutions similar to PLM systems to achieve support for information integration. But, the nature of AEC projects requires a more pragmatic approach for PLM systems to be implementable. Modularisation concepts in standards and frameworks for information management – such as the IDM and LOD solutions in Paper A, B and C – address this challenge. Instead of needing to define the PLM structure and relations in each unique AEC project, product information requirements and relations to the process model can be automatically generated based on predefined standard work packages using e.g. the proposed IDM Manager.
Such solution requires initial work by national organisations, suppliers, clients, consultants, etc. to predefine the content of work package libraries. Considerable work is already ongoing in many countries to defining product information requirements, but so far the work is difficult to harmonize and reuse. An agreement to use a standard framework for IDM and LOD development – as proposed in this thesis – could not only improve standardization efforts but it could also provide a basis for improved abilities to introduce PLM systems to the AEC industry. The modular yet structured approach allows for automation of information requirements, flow optimization and identification of relations between information models, lowering the barriers for PLM system implementation.
The main focus in this research has been related to the process and the product model, but it seems that a similar modular approach could be applicable for linking the other information models as well.
Creating a library of mission deliverables for the mission model could for example include derived functional requirements. The Danish energy frame requirements (The Danish Transport and Construction Agency 2015) with ‘Building Class 2015’ and ‘Building Class 2020’ is one example which allow clients to describe their ambition for energy performance of a building by simply referring to a certain energy class. This class translates into a set of functional requirements for energy
Part IV – Discussion and Conclusion
Digitalization as Driver for Standardized Specification and Design of Buildings 57 consumption, window performance and ratio etc. – generating automatic content for the function model.
A specification defining the link between the functional model and the product models is already established and described by (Kiviniemi 2005). Based on this specification, the relation between requirements in the function model and objects in the product model can be established automatically for alignment between the two models.
Based on the type of building systems defined in the function model, it would also be possible to generate a work breakdown structure for the work to complete in the process model. For example, if cooling requirements are specified, engineering work to dimension a cooling system is required, and if vacuum outlets are required, a utility engineer dimensioning a vacuum system is required.
Relations between the functional model and tasks in the process model can in this way also be built automatically if prior concepts for the relations between systems and activities are defined.
The modular approach adds value when used in relation to activities and processes in building design which are repeated across projects. Paper B illustrates that there is considerable overlap in work packages in the two different case study projects with almost 70 % of the work packages being reusable. A modular approach should, however, not aim to represent all possible project and organization types. Yet, using predefined packages can provide a consistent starting point for building content in a PLM system, and this seems highly attractive in order for such systems to be implemented in the AEC industry.
Combining the above findings – in particular the Scrum based APM methodology proposed in Paper E – with the combined information model described in Section 7.1, a methodology for building design management is proposed as shown in Figure 25. The methodology is referred to as the mission, function, process and product (MFPP) methodology for building design management and describes how to integrate information, organization, process and product into a platform which supports an agile development process in a pragmatic approach.
The client brief was included in the illustration of the combined information model in Section 7.1, but is removed in the MFPP methodology. Instead this author argues that the client brief should be an integrated part of the methodology – a starting point for the first content in the information models. This will ensure far better integration between client needs and design solutions and motivate further for keeping the mission model continuously updated throughout the design process.
Part IV – Discussion and Conclusion
58 Digitalization as Driver for Standardized Specification and Design of Buildings
Figure 25. The MFPP methodology proposed as a result of the research in this thesis
In brackets in the illustration above is the APM terminology used in Paper E to illustrate how the information models relate to the APM development steps. The illustration in Figure 26 is introduced in Section 6.4 and also represents the APM development steps but in general terms. The map in Figure 26 can be used as a guide to understand the intention with the MFPP methodology as following:
1. The Mission model should capture information from the requirement analysis step which is referred to as story development in APM.
2. The Function model should capture information from the functional analysis step which is referred to as feature development in APM.
3. The Product model should capture information from the synthesis step which is referred to as the design solution developed incrementally during sprints in APM.
4. The Process model should capture information on tasks required in primarily the requirements and design loop referred to as backlog prioritization and sprint planning in APM.
Project Req.
Building Req.
Site Req.
Storey Req.
Space Req.
Building Envelope Req.
Circulation System Req.
StructuralSystem Req.
HvacSystem Req.
...
TelecomSystem Req.
FireSafetySystem Req.
Project
Building Site
Storey
Space
Bounding
Elements Surface
Materials
Furniture Equipment
Building Envelope Circulation System Structural
System HVAC System Telecom System
...
Security System FireSafety
System ...
Project Mission
Sub-mission (Epic) Sub-mission
(Epic)
Delivery (Story) Delivery (Story) Delivery (Story) Delivery (Story)
Delivery (Story) Delivery (Story) Delivery
(Story) Delivery
(Story)
Sub-mission (Epic) Sub-mission
(Epic) Sub-mission
(Epic)
Function Model
(Feature Development) Product Model
(Design Solution) Mission Model
(Story Development)
Objectives Short-term value Long-term value
Sprint 1
Architecture Interior arch.
Facade arch.
Landscape arch.
Structures Concrete
Steel Building Services
HVAC Piping Electrical
Package 1
Package 3
Package 2 Package 4
Package 5
Package 6
Package 7 Package 8
Package 9 Package 11 Package 10
Package 12 Package 13
Package 14 Package 15 Package 16
Process Model (Backlog Prioritization)
Work Breakdown Task Sequencing
Disciplines
...
Clarification 1
Clarification 2
Clarification 4
Clarification 3
Clarification 5
Clarification 6
Clarification 7
Change 2 Task 4
Task 2 Task 7
Task 3 Change 1
Change 3 Task 7
Task 3 Task 7
Task 3
Task 7
Sprint 1 Sprint 2 Sprint 3 Sprint ...
Package 1 Package 2
Clarification 1
Clarification 2 Task 4
Task 2 Change 1
Improvement Verification
(Retrospective) Additional tasks
The MFPP Methodology for Building Design Management The MFPP Methodology for Building Design Management
(Sprint Review) (Sprint Planning and Execution)
Sprint 2
Package 3 Package 4
Clarification 4
Clarification 3 Change 2 Task 7 Task 3
Sprint ...
Package X Package X
Clarification X
Clarification X Change X Task X Task X
Part IV – Discussion and Conclusion
Digitalization as Driver for Standardized Specification and Design of Buildings 59
Figure 26. Map of the design process from both a systems and a design thinking perspective – identical to Figure 12
The process model should be considered as a combination of the task management platform proposed in Paper F and the IDM Manager presented in Paper B. Some distinction is required in this regard as the term ‘task’ is used differently in the scientific papers. A task should be considered as a general term for any elements in a process. In Paper E, four types of tasks are identified as design issues, tasks (here as low-level tasks), clarification issues and change requests. Design issues should, however, be split into deliverables, which are defined in the Mission model, and work packages which are defined in the Process model. This differentiation was a lack in the APM methodology tested because it mixed what should have been stories of client needs with high level design activities – probably adding to the difficulties of describing clear functional requirements in the case studies. Work packages for the design activities should be used to plan the information flow efficiently in the IDM Manager and all additional tasks related to decision-making, feedback etc.
should be planned alongside the work packages to prioritize which tasks are to be solved in what sprints. Tasks related to decision-making and feedback should not be included in the IDM Manager, as these should not influence the information flow planning as discussed in Paper B, page 10.
The MFPP methodology is not tested within this thesis as it is proposed retrospectively to the work completed. However, several elements have been evaluated in the scientific papers resulting in findings which support the content of the MFPP methodology. For this reason, the methodology is believed by this author to be a valid solution for a pragmatic approach to a PLM platform to the AEC industry which supports the needed agile development steps. As discussed previously in this section, it is further supportive of modular approaches which allow for automated creation of internal structures and content along with creation of relations between the information models.
This research, furthermore, makes use of a range of different theories and methods which have previously been evaluated individually in the AEC industry and found useful. Based on the findings in this thesis it is clear that a multitude of theories and methods are required to support the diverse needs in building design management. The existing theories and methods should for this reason not be considered as alternatives to each other but as elements in an integrated approach. A key challenge ahead for the AEC industry is to find ways to integrate these theories and methods rather than executing these in parallel and thereby not achieving the required level of improvement. The MFPP methodology can serve as a contribution to how several perspectives can be integrated in a common approach for efficient building design management.
Functional
Analysis Synthesis
Requirements Analysis
Decisions (Design Output) Mission or
Problem
(Design Input) Requirements
Loop
Design Loop
Verification Loop
Evaluation Learning
Loop
Improvement Loop Design Thinking Perspective
Systems Engineering Perspective
Part IV – Discussion and Conclusion
60 Digitalization as Driver for Standardized Specification and Design of Buildings