An approach to modelling the dynamics of the design process in architecture

This study presents a brainstorming on the dynamics of the design process in architecture. It is motivated by the needs and the evolution of the professional practice, and based on their study; it is exposed to allow us to learn the appropriate lessons for a future modeling of this dynamics. We will introduce our working hypothesis of the component dynamics and suggest a model for dynamics of design process in architecture. 2.1 The collective and shared natures of the design from Co-design to distributed design. In practice, most of design activities combine various contributions made by different actors. In the construction industry, Callon (1997) displayed this collective and negotiated nature. With the architect, responsible for the architectural design, several engineers specialized in various technical fields (e.g. structure, acoustics, thermics...) take part in the design of the building, on the basis of specifications defined by the client. The execution studies that consist of setting up methods of building construction and are carried out by the contractors, are becoming more and more important in the building design. As described by (Darses & Falzon 1996, Turk et al 1997, Hanser 2003) the implication of actors in a design process can take various forms. Their engagement in the process is similar to a Co-design or a distributed design (Figure 1). The actors can meet these two situations successively, during the same project or the same design process. Thus, at engaging meetings or at definition of requirements, it is in the dynamics of Co-design that makes the whole group work together. The actors develop a joint solution. They share an identical goal, which is to define a requirement of solution to achieve the expressed objective. All these actors use specific skills but with strong collective constraint to attain a common goal: the choice of a requirement of solution. The choice of principles marks the beginning of structuring the future works, the actors are no more jointly but simultaneously implied into the process. Then they are going to work on specific tasks, related to their specific skills. We are close to the concept of the distributed design here. These actors strive towards their own aims and objectives, while taking part as effectively as possible in the collective definition of work. Figure 1. Distributed design and points of synthesis. Hanser, (2003) according to (Turk et Al. 1997) 2.2Design process from divergent angles The design process has been researched by various scientists from different points of view. On having carried out a deep analysis of their works, we would like to present the results they have received under four dimensions: communication, organization having an influence on communication, cognitive aspect of design process and planning of design activity. First, it is important to notice that many scientists pay our attention to the complex nature of the dynamics in the design process and underline the existence of various reasons that have their influence on it. However, they do not define composing dynamics and leave the question of their formalizing open to discussions. It is also necessary to add that the analysis of the dynamics of the design process is rarely a key factor of the works we have analyzed. While presenting a model, the authors mark its importance; still the implication of the proposed models can be hardly provided “in pertinent changes of distributed environment” Chiu (2002). The role of organisation in design communication and collaboration is studied in “An organizational view of design communication in design collaboration” Chiu (2002). The author presents a process model of design collaboration that is supposed to be facilitated by a structural organization and a computer-supported collaborative work. Having a project-oriented character, the architectural design practice is well described by design organisation and design activities. While speaking on communication problems in design collaboration, the scientist underlines the dynamic character of media, semantic, performance and organizational aspects and their influence on design collaboration of the whole project. The particular attention is put to the information flows in design communication, the participation of individuals in each task and the coordination of design information. The author also tends to illustrate his approach by case studies and design experiments. As for the cognitive aspects of the model of Chiu, the article gives a basic understanding of the design collaboration in the architectural practice and formulates the main principles of design strategies. The author underlines the importance of three levels of communication (individual, group, project) but he leaves a reader to formalise differences between these levels as well as to seek supplementary mechanisms for the direct application of one model in the design process in architecture. The interdependence between the communication factor and the cognitive approach in design is presented by Stempfle & Badke-Schaub (2002). The authors introduce four basic cognitive operations of designed research that establish a generic model of design team activity – generation, exploration, comparison and selection. They focus on the analysis of main cognitive operations of the design process – generation and exploration, which are followed by mechanisms of its natural evolution – blind variation and selective retention. Methods of comparison and selection are chosen to narrow the problem space and to detail the basic elements of thinking in design teams. The proposed approach enables the decomposition of the team-design process that gives the possibility for planning in the terms of contentand process-directed activity. The authors propose a parallel analysis of the activities directed towards the content of a design problem and of those directed towards the organisation of the group process. The general character of the model is explained by the chosen analysis (analyse of frequencies, process analysis under a macro/micro perspective). This scientific viewpoint limits its industrial implementation: “design methodology has not been as readily accepted in industry as design methodologists have expected”. It is also important that the model is not particularly made for architecture, but it seems to be rather flexible for modelling interactions which form the base of the dynamics. The cognitive approach is well described in the other article by Chiu (2002). The article proposes a descriptive model of the situated design detailed by a design experiment, case adaptation and comparative analysis of the design situations. A research paradigm of case-based reasoning and main principles of design moves are taken into account in the analysis of the designed situatedness – the situated activity of the interaction among “designers, cases, programmes, individual workplaces and tools” Chiu, (2002). The observations and the analysis of their results are proved by a design experiment from the point of view of various participants. We find it essential to underline that the model describes low-level interactions. That is why the learning ability of situation identification and case adaptation of the model could be hardly applied for the complex process of design in architecture. Mao-Lin Chiu also proposes to apply constructive memory of cases and design tools of the model for future development planning of suitable computational environment for designers. The model establishes the relation between the behaviour of designers in new situations and the routine design process. On the other hand, the metrics based analysis model (Xijuan et al. 2002), reflects the general approach of the design process. The scientists propose a mathematical model of process planning that provides high design performance within a limited schedule and budget. They conduct an effective task analysis on optimizing the process of design organization measured by different degrees of task importance. Starting by task decomposition and computing the critical degree of each task, the authors receive an evaluation of the likelihood of error occurrence for each task-contributing factor. The total degree of the task importance is found by adding its spread degree that reflects the interdependence of all factors. Therefore, the task analysis model forms a base of the design process planning followed by the evaluation of results. The classification of the factors into critical and non-critical is made after the cognitive analysis of a full-scale and open-architect database of design process systems. However, this subjectivity and lack of the possibilities to manage and control the design process turn out to be disadvantages of the approach. Speaking on the design in architecture, we require a more flexible model that would represent the dynamics of the process and illustrate its “architectural” characteristics. 2.3 Conclusion of the state of art In short, previous studies have focused on issues of design collaboration including the process, the team works, the design settings groupware, the organization teams; and they influence on communication patterns. However, these studies rarely focus on the importance of the whole dynamics of design process in architecture. The following sections will present our brainstorming on the dynamics of the design process in architecture and our model suggestion. 3SUGGESTED APPROACH AND SCIENTIFIC STEP OF DEVELOPMENT In fact, social, temporal, and cognitive contexts play a considerable role in the evolution of an architectural project. These contexts cannot be taken into account in such a structured approach of the design. Therefore, it is important to consider the design in architecture not only as an activity of resolution of issues, but also as a complex activity, that has to reckon technical, temporal, social, strategic and economical aspects at the same time. The design process in architecture can be compared to “a living being” having continuous input flows (e.g. adding a document, arrival of a new actor, creation of a project ...) and output flows (e.g. end of a task, deletion of a document...). So, we notice the existence of input and output data which have to be defined and validated. In that way we progress the design process in its every identified stage. That allows us to speak on a certain generation and destruction of the components that keep this process in a stable state. Such stability is one of the main criteria of the design process as it guarantees its proper functioning. However, the design process reacts to the temporary changes of the environment and to “stimulus” by the reversible fluctuations of its stable state (e.g. taken into account new legal requirements, administrative and judicial procedures, technical requirements ...). These changes can be regarded as temporary perturbations which convert a “being” into “equilibrium” at the flow equal to the stable state. The multivarious intern and extern interactions and the dynamic organization of the design process in architecture underline its systemic character; so it turns out to be an open system having a goal of constructability of the architectural project in design. We can also notice that the design process in architecture is an ordered system of sub-systems in interaction. These sub-systems also constitute a certain order: the represented state on each observation level is just a result of exchanges between the components of lower levels (Figure 2). This order is not a resultant of rigid functionalities of the components; it results essentially the coordinating interactions among the components which roles are not constant. So, according to its internal and external multivariable interactions and its dynamic organization, the design process in architecture can be compared to a measurable open system. Figure 2. Systemic diagram of the design process in architec-