This paper proposes a strategic framework for how computational design can be transitioned and upscaled from technical development to full industrial project implementation in live commissions, providing opportunities for innovation within the economical and time-based constraints of the architectural commission. It primarily addresses computational design development as part of ongoing commissions and identifies three main challenges in how this can provide opportunities for project specific innovation. The first challenge regards the scope and expected benefits of computational design development in relation to the overarching scope of a commission in situations when the project lead is unfamiliar with the methodology. The second challenge is how to manage the development process in relation to the overarching process of the architectural commission. The third challenge regards the interaction between developers and project architects/engineers in terms of expectations on roles, responsibilities, and mandates, something which is also affected by the selected workflow model. The responses to these challenges are based on the experience of the 13-year practice of Dsearch at the architectural practice White Arkitekter, and involves the definition of new offers, models for alternate workflows, and a mapping of developer roles and responsibilities in four dimensions. Three design cases are used for clarification and reflection, each with its own condition and scope.

This paper presents the design development of The Sheaf - a 12-meter-high iconic timber-diagrid hyperboloid lookout tower, from concept to detailed design. The Sheaf was the winning proposal in an architectural competition in 2019 for a lookout tower in Varberg, Sweden. A close dialogue between architect and engineer propelled the iterative design development, initially exploring the concept and later solving details leading to a sustainable, material-efficient, and expressive tower using simple, affordable building elements. The alternating use of digital parametric models and physical model tests supported the decision-making, allowing quick exploration of viable alternatives and the client to get involved.

This paper presents two cases of material reuse in architectural practice, employing bespoke computational workflows. The first unpacks a complete cycle of from building demolition of an existing building in Järvsö, Sweden, through the design and fabrication of a temporary and mobile pavilion for use at a series of urban events, to its final destination as a wind shelter not far from the forests where the material was initially harvested. The second presents scaled approaches and methods to respond to the competition brief for the redevelopment of a partial urban block in Berlin, using the material stock from the existing building as a material library. Both cases have been developed in constrained project contexts where the resources of the material libraries have been limited and directly associated to the project - rather than being part of an open market of re-used resources - allowing data on availability, amount, and quality to be readily available. In this sense the approach can be defined as a project-confined re-use workflow, providing the opportunity to target the association between the design modelling environment and the material library. 

This chapter describes a research thread at CITA which explores how computation and a challenging of traditional material practice can impact the use of timber in architectural design and fabrication. Several past research projects at CITA have demonstrated the potential for streamlining the design-to-production process using computational tools, and the value of working in concert with the inherent properties of wood. Current research continues this thread through a participation in the Innochain research network (http://innochain.net/ ) and collaboration with industrial partners White Arkitekter AB and Blumer-Lehmann AG. Through the embedding of digital tools within established timber design a fabrication processes, new workflows are proposed which could lead to more intelligent design decisions, optimized building components, and new timber morphologies. 

In this paper we address the problem of knowledge management in architectural design computation practice, reflecting on our practice at Dsearch-a design computation network within White arkitekter. As a means to investigate relevant aspects of visual scripting, we introduce the notions of code, algorithm and note. We also introduce two different modes of operation within architectural practice: morphology and development-which help us distinguish the diverse knowledge types typically occurring in the structure of visual scripts. We describe two sets of tools developed by Dsearch to continuously integrate planning and documentation with design development work. The main conclusion from our practical experience of this approach is that it allows critical reflection into an efficient workflow. This constitutes a new kind of practice based and action oriented knowledge that can be curated in the form of design narratives.