Prefabrication 4.0: BIM-aided design of sustainable DIY-oriented houses

Abstract

In the context of continuous housing shortage, increasing construction standards and rising labour costs, one of the possibilities to address this array of problems is prefabrication directed towards do-it-yourself (DIY) construction methods. This paper presents a prototype tool for aiding the design of DIY-oriented single-family houses with the use of small-element timber prefabrication. The introduced solution uses the potential of BIM technology for adapting a traditionally designed house to the prefabrication requirements and reduction of waste generated in the assembly process. The experimental tool was developed in the Autodesk Revit software. It incorporates custom Dynamo-for-Revit scripts. The experimental tool implemented the user- and technology-specified boundary conditions and converted an input BIM model into a prefabricated alternative. The tool was tested on the design of a two-story single-family house. The results compare the automated optimized panelization with manual approach. The simulation revealed the possibility of the construction waste reduction by at least 50% when comparing to the non-optimized panelization.

Keywords

DIY construction prefabricated house timber prefabrication small-panel prefabrication BIM-aided panelization Building Information Modelling

Introduction

The Fourth Industrial Revolution blurs the boundaries between the physical and the digital. The impact of the Fourth Industrial Revolution on architectural design is tied to the development of CAD/CAM/CAE. Contemporary digital tools and techniques not only speed up the architect’s work, but also contribute to integration of disciplines, enable design simulations, automation of both design and manufacturing processes, and creating solutions tailored to designer’s individual needs. This digital paradigm gains importance in residential design. In the face of the current housing shortage combined with rising labour costs in developed countries, shrinking resources, changing climate and evolving social needs, the housing industry requires a change of approach.

Experience gained from the 20th century shows that reducing construction time and costs as well as lowering rents is tied to standardization of architectural solutions, usually associated with prefabrication.¹ Historically, a prefabricated building was constructed with limited number of factory-made components.² A small variety of used components resulted in highly repetitive typified architecture.³ Examples of such repetitiveness are prefabricated concrete multifamily housing erected in the mid-20th century that are still present in the cityscapes in the Central and Eastern Europe.⁴ According to Tatjana Schneider and Jeremy Till,⁵ in most cases this kind of prefabrication proved to be ineffective in terms of building flexibility, which is commonly referred to as a determinant of a well-designed housing. A house or apartment should evolve along with the users’ evolving needs. The authors argue that most of the known prefabricated housing of the previous century fail to meet this requirement.

Schneider’s and Till’s view converges with the philosophy of Walter Segal. Segal is known for his self-build method which extends the Schneider’s and Till’s idea of flexibility by favouring a do-it-yourself (DIY) approach.⁶ The DIY approach enables the unskilled users to construct houses of their own and to adjust them according to their changing needs. Segal’s method assumed prefabrication using simple materials such as timber. He opted for using standardized elements rather than inventing a new standardized system. By promoting prefabrication and eliminating qualified workers from the construction site, he aimed to lower the construction cost, and therefore, increase the housing affordability. At the same time, the term ‘flexibility’ can be applied to the material itself,⁷ as standard timber elements, if required, can be processed manually on-site in order to achieve the desired form.

In recent years, prefabrication technologies improved significantly compared to the previous decades. As claimed by Nikos Salingaros and Débora Tejada,⁸ a valid alternative to modular design is determining a building structure through subdivision of the architectural form. According to the authors it enables the structural components to assume any shapes and sizes, liberating the architecture and its users from the limitations of a rigid modularity. Thomas Bock and Thomas Linner⁹ argue that development of CNC manufacturing decreased the importance of modularity of the prefabricated structures. They assert that the mass customization paradigm allowed for designing architecture adapted to individual needs on an industrial scale.

Nowadays, flexibility and affordability are still challenges in housing design. This creates the opportunity for architects and engineers to take advantage of technological advancements of the Fourth Industrial Revolution to develop solutions addressing these challenges. One possible option is to incorporate Segal’s and Salingaros’ ideas into an integrated tool supporting the design of flexible timber structures suitable for prefabrication. In the last few years timber prefabrication has significantly gained in popularity. Timber offers a wide range of applications: from large scale panels, such as cross-laminated timber (CLT) to timber frame structures. In each application there are many opportunities and potential for digitally enhanced design and construction.

Computationally aided timber prefabrication was explored by Mehmet Sinan Bermek et al.¹⁰ The authors combine Building Information Modelling (BIM) methodology with graph theory for enhanced structural design. Their goal was to develop ‘processing engines, schemata and active databases that can be used in automated decision making, generative design, and dynamic optimization while maintaining human readability’. Their focus, however, was limited to large span planar elements such as CLT components, which is not conducive to the DIY approach.

Oliver Neumann and Daniel Schmidt¹¹ conducted research on sustainable CNC production of timber framing. They not only explored the resource and material efficiency and the design and simulation of the prefabricated structures with digital models, but also reflected on ‘structural performance of timber framing methods and traditional wood-to-wood joints that were allowed to thrive due to the abundance of large timber in smaller square sections’. Joinery itself is also a frequent subject of research,^12,13 however, this is a separate subject valuable for another paper.

Offsite timber construction was researched also by Panagiotis Patlakas, Andrew Livingstone and Robert Hairstans.¹⁴ The authors proved the potential of BIM which not only integrates architectural and structural technology, but also offers tools for analysing life cycle aspects and performing other simulations. However, they point out that ‘the major implementation challenge comes mostly from the integration with the manufacturing process’. Patlakas et al. argue that the industry should establish a common BIM collaboration standard.

The available publications reveal a research gap in the field of digitally aided design of DIY-oriented prefabricated timber houses. Apart from the aforementioned examples, contemporary research in that field is dominated by the exploration of concrete precasts^15–17 and concern mainly CAD-aided fabrication methods such as computational generation of moulds and formwork.^18,19 However, multiple studies^10,14–17 have indicated usefulness of BIM in design with prefabrication. BIM environment can be successfully used as a platform for simulating the prefabricated structure.^15,17 This prompted me to undertake further research in this area.

The flexibility and affordability challenges could be addressed by introducing a small-element timber prefabrication. Panels made of a single layer of solid timber lamellas (Figure 1) can benefit from all the advantages of digitally aided manufacturing.²⁰ Because of limited sizes they can be handled manually by 2–3 people. The solid timber structure allows for cutting, milling and drilling in any desired direction. The panels can be joined with nails and screws. Due to simple carpentry joints the partition walls can be demounted and placed elsewhere if needed, increasing the structure’s flexibility and its DIY potential. However, the process of timber construction, as in other building technologies, generates waste in the form of unusable cut-off elements (Figure 2). In times of growing environmental awareness, it is justified to seek for solutions that balance flexibility, construction costs and waste reduction.

Figure 1.

Assembly of small timber panel house. Standard glulam panels of size 260 cm × 60 cm × 10 cm designed to be assembled by three people. Design: Svobo (www.svobo.pl). Photo: courtesy of Artur Stępniak.¹⁷

Figure 2.

A schematic plan of panels join at wall corner.

The study presented in this paper aimed to bridge the flexibility of timber prefabrication with DIY and minimization of timber waste. It extends the research demonstrated at the eCAADe + SIGraDi 2019 Conference.²¹ It covered a BIM-based solution which supports the design of timber prefabricated single-family houses. In reference to Salingaros’ and Tejada’s view, the assumed approach intends to enable creating a model using standard tools available in the specialized BIM software, such as modelling composite walls and placing predesigned components (e.g. windows and doors). The presented solution utilizes data stored in the model to simulate the small-element prefabricated panel structure suitable for DIY assembly and to optimize the panelization in terms of generated waste. In summary, the purpose was to comply with the mass customization paradigm by not limiting creative freedom to the set of predesigned standardized building components.

Methods

The process of implementing a small-element prefabrication into the design follows a set of rules that can be formalized, and therefore, become automated. The presented solution uses a typical method of adapting the existing design to the specifics of small-element timber prefabrication. To assure the automation of the conversion, the solution combines BIM with generative tools and custom scripts. The general workflow (Figure 3) assumes collecting the input data, such as prefabrication technology constraints and other design parameters, and combining them with the original BIM model in the analysis phase. After the analysis, the splitting and optimization algorithms generate the split pattern and transform the model into the prefabricated alternative. The prototype was created in the Autodesk Revit environment extended with custom scripts developed in Dynamo-for-Revit extension with the use of Revit Application Programming Interface (Revit API).

Figure 3.

The prototype tool process workflow.

Prefabrication technology constraints

As any other building technology, timber prefabrication also comes with technological constraints. These technological boundary conditions need to be provided by the designer before the conversion process. On this basis the algorithm computes the maximum sizes of prefabricated elements. During the conversion process the tool acquires the necessary data from the BIM model as explained in the Model Analysis – subsection to follow. For walls, the maximum panel dimensions are determined based on the building net level height H , expected structural wall thickness, maximum panel weight defined by the user, and average structural wood density. The weight value can be specified based on the local labour law or users’ individual preferences. During assembly the panels are joined with grooves milled along each vertical edge (Figure 4). The effective panel coverage is reduced by the width of the grooves - approximately 15 mm.²⁰ Due to the manufacturer’s fabrication constraints, the panel width in centimetres should be a multiple of 4 and should not be smaller than 36 cm. The computed values are used as boundary conditions for panelization. In this paper the panels of maximum allowed width will be further referred to as standard panels and their width will be marked as w* .

Figure 4.

A schematic horizontal section of prefabricated small timber panels for demonstrating the absolute panel width compared to its effective size in the assembled structure.

The small-element timber prefabrication requires introducing additional constraints. To assure a proper structural stiffness each wall needs a pair of foundation and cap beams of [structural layer width] × 10 cm profile, which together decrease the standard panel height h by 20 cm. Similarly, a similar beam needs to be applied for each windowsill. Furthermore, like in traditional construction, both windows and doors require lintels which span over panels directly neighbouring the opening.⁷ A typical subdivision of a wall is shown in Figure 5.

Figure 5.

Example of wall subdivision including all key structural elements and its sizes. For panels of w* = 40 cm the coverage width is 38.5 cm.¹⁷

Design parameters

The presented tool also involves design-specific parameters, which are not directly related to the prefabrication constraints but rather to the building’s geometry and designer’s intent. For example, for structural reasons the designer must make sure that a window and a door lintel will have sufficient support. Therefore, for each opening a lintel margin m₀ needs to be provided, ensuring the splitting which occurs close to the opening would not decrease the lintel’s structural performance. The value of the lintel margin should be coordinated with the structural engineer. The operation of a lintel margin is explained in the Splitting and Panelization subsection.

Model analysis

In the first phase the prototype analyses a standard BIM model from which it extracts information about the building geometry. The model geometry is examined on three levels:

Solid geometry – containing information about objects’ dimensions and their orientation in the model space.

Lines – a parametric reference for linear elements, such as walls.

Points – references for the hosted elements locations, such as windows and doors.

The algorithm examines the model to obtain its geometric structure. First, for each composite wall the script extracts the structural layer and its thickness t (Figure 6(a)) that is used for calculating the sizes of prefabricated panels. The input model is then simplified to lines (wall centrelines), points (locations of openings) (Figure 6(b)) and correlated data such as windows’ and doors dimensions. Afterwards, the script sorts wall location lines into chains. This procedure examines the lines’ endpoints. The overlapping ends indicate the lines’ neighbourhood and such lines are appended in the right order to collections representing each wall chain. The sorting is performed to assure a logical organization of data in the simulation and in the final documentation and assembly instruction.

Figure 6.

Model analysis, splitting and panelization workflow for demonstrating subsequent simulation steps. (a) Extracting the structural layer. (b) Determining model elements’ locations. (c) Establishing model hierarchy. (d) Projecting opening boundaries on the walls’ location lines. (e) Generating preliminary division scheme. (f) Generating structural beams for walls and openings.

The extracted data is then organized hierarchically. Due to the specificity of the chosen software the retrieved data structure reflects the Revit model hierarchy: the elements are organized with respect to the building levels, category, host (for openings) and type (Figure 6(c)). However, similar data structure could be used universally. The script creates data collections for each kind of model elements: separate for walls, slabs, windows, doors and other openings (if applicable). Each opening is associated with its host geometry, for example, each window is linked to the wall in which it is placed. Furthermore, windows and doors are assigned with types, such as ‘Door D1’, ‘Window W1’, and so forth. All elements must be correctly defined in a BIM model as they are automatically processed in Dynamo.

Splitting and panelization

In the second phase, the openings are projected onto location lines indicating the span of each opening in its host wall (Figure 6(d)). This information is used to determine the sizes of window cap beams as well as sizes of lintels for all openings. Afterwards, the splitting routine generates a preliminary division scheme containing a collection of split points on the wall location lines (Figure 6(e)). The preliminary division is generated for each wall by sequential adding of split points, starting from each wall’s ending. The location of split points fits the prefabrication constraints, as mentioned previously. Due to these constraints, the division might result in several non-standard panels.

Subsequently, the pattern of wall foundations and caps, as well as opening caps and lintels is added to the division scheme (Figure 6(f)). Wall foundations and caps are generated along all walls in the project. Window caps are created based on windows’ widths. Opening lintels are computed with the additional implementation of the splitting margin, as described in the Design Parameters subsection. If the splitting point occurs outside of the opening’s safety margins, the lintel will span over panels directly neighbouring the opening. If the splitting occurs within the safety margin, the lintel is extended over the next neighbouring panel (Figure 7).

Figure 7.

Resultant shape of an opening lintel depending on the generated division scheme and the size of the user-defined opening safety margin.

In the following phase the division scheme can be further optimized to minimize fabrication cost and construction waste.

Optimization

In buildings not designed with modularity in mind a perfectly repeatable panelization is very unlikely to occur. Most often the resulting structure includes several unique elements. These elements usually appear in walls’ endings and near openings. These custom panels can be fabricated off-site with CNC machining or by manual on-site adjustment of the standard ones. In both cases the cut-off fragments of the panels are a potential waste. The additional aim of the developed tool was also to minimize the amount of waste generated in the process. In the proposed solution, the panelization is generated in such a way that as many unique elements as possible can be combined into larger standard panels. In other words, it is intended that the leftovers, such as cut-offs appearing at wall endings, can be reused elsewhere in the structure.

The algorithm adds all the irregular panels to the collection E :

E = {e_{i} | w_{i} < w^{*}},

where w_i denotes the length of the panel e_i . Next, it pairs them in such a way that each two panels form a combination as close as possible to the size of a standard panel. A separate collection E₀ is created for panels that cannot be paired with any other element. Afterwards, the script calculates the width of the leftover from each pair:

l_{i} = w * - (w_{i 1} + w_{i 2}),

where w_ik stands for the width of each ending panel. For the panels from the collection E₀ the leftovers are calculated as:

l_{i} = w * - w_{i},

The summary volume of these leftovers is marked as L_e .

L_{e} = \sum_{i} l_{i} \cdot h_{i} \cdot t_{i},

where h_i stands for the height, and t_i stands for the thickness of the panel e_i .

In parallel, the prototype calculates leftovers from other atypical panels that appear near openings. The leftovers o_i from these panels are obtained by cutting the shape of the opening or by shortening the panel for the purpose of the lintel as shown previously on Figure 5. These leftovers cannot be adapted elsewhere in the structure. The summary volume of these leftovers is marked as O .

O = \sum_{i} o_{i},

The volume of all the leftovers is then summed in order to indicate the total amount of waste:

L_{t} = L_{e} + O .

The optimization routine was intended to balance the following three criteria:

Maximum panel width: the algorithm favours the highest width obtainable with the user-defined weight constraint. This criterion is related to the total panel count. It is assumed that the lower total panel count is tied to the shorter construction time.

Minimum waste: the algorithm favours the panelization where the most leftover panels are used in other parts of the structure (minimizing the L_e value).

Minimum type count: the algorithm favours the solutions with the most repeatable panelization. This criterion is related to construction cost as all the irregular panels require additional work as they are obtained either by additional CNC machining or by manual adjustment on site.

The optimization process is performed iteratively. The algorithm choses the pair which generates the biggest leftover and shifts the panelization in the wall with the shorter element of the pair so that this element absorbs the leftover. This operation generates an irregular panel on the other end of the wall. This panel is then appended to the collection E₀ . Subsequently, the script updates the L_t value and initiates another iteration. The process continues until eliminating all the leftovers l_i .

Next the process is repeated for panelization with

w_{i}^{*} \in {w_{i}^{*} | w_{i}^{*} = w^{*} - 4 n, n = 1, 2, \dots \land w_{i}^{*} \geq 36 c m} .

Finally, the script returns the solution space, which can be examined based on the criteria mentioned previously.

Geometry processing and documentation

The final panelised geometry is achieved by creating Revit parts for each structural component. The parts stand for wall layers and are a feature specific to Autodesk Revit environment. These parts are then split based on the division scheme. Afterwards, each panel is assigned with a unique id code necessary for on-site assembly. And finally, for each building wall the program generates a shop drawing and detailed schedule for fabrication and assembly.

Results

The presented simulation tool was tested on the design of a two-story single-family house with an area of 80 m², story height without finishing of 271 cm and external walls perimeter of 32.24 m (Figure 8). The developed algorithmic solution was intended to introduce the benefits of industrial advancements to small scale realizations. The converted model was prepared in the Autodesk Revit software with the use of standard modelling tools. The composite walls had properly defined layers so that the algorithm was able to extract the structural layer for further processing.

Figure 8.

The analyzed building scheme including basic dimensions. From the left: ground floor plan, first floor plan, cross-section.¹⁷

With the use of small timber panels, the structural wall thickness could be reduced to 10 cm. This allowed for saving approximately 3.2 m² of space in comparison to traditional brick technologies. To assure the proper structure stiffness, each wall required a foundation and cap beam of a 10 × 10 cm profile. As a result, the effective height of a prefabricated panel was 251 cm. The analysed building requires a total of 25.27 m³ of solid timber for all structural components. For panel size calculation the average wood density was set to 450 kg*m^-3. To enable the assembly to be pursued by two people the maximum panel weight was set to 50 kg. Based on these constraints, the algorithm calculated the maximum panel width to be 44.26 cm. To meet the fabrication constraint of width being a multiple of 4 cm this value was then rounded down to 44 cm. The resultant standard panel measured 44 × 251 × 10 cm.

In a manual division process, the panelization is established for each wall separately, starting from an arbitrarily chosen end. This approach produces the highest waste rate, since usually the atypical ending panels are prepared from the standard panels on site and the cut-off parts are not reused. For the standard panel division, which was performed analogically to manual panelization, the non-optimal total panel count was 282 (Figure 9(a)). The outcome included 146 standard panels, 14 atypical panels requiring manual or CNC adjustment due to openings in walls, and 26 unique panels at wall endings. Only four of these unique panels could be merged into standard-sized elements. Moreover, there were also six shorter panels under windows, that could be merged into standard panels and 20 panels neighbouring door openings which were shortened due to the appearance of opening lintels. Similarly, 78 shorter panels appear also in the attic level. These attic panels, however, can be produced in their final size during fabrication. For the standard non-optimized panelization the construction would result in 26 waste leftovers of total volume 1.05 m3 which stands for 4.2% of the timber required for the construction of the analysed house.

Figure 9.

Comparison of panelization results for demonstrating the outcomes achieved for different standard panel sizes.¹⁷ (a) Preliminary panelization for the 44-cm panels. (b) Optimized panelization for the 44-cm panels. (c) Optimized panelization for the 40-cm panels. (d) Optimized panelization for the 36-cm panels.

Afterwards, the algorithm performed the optimization procedure in order to improve the achieved panelization. The optimized distribution of standard panel division (Figure 9(b)) featured a better adaptation of the cut-off fragments. It allowed for reduction of the waste generated by cut-off fragments by 0.55 m3 (52%) while the panel count remained the same. Next the algorithm performed the simulation with standard panel width decreased to 40 cm and to 36 cm. These steps improved the result in terms of generated waste by 70% and 64% respectively in comparison to the non-optimized outcome (Figure 9(c) and (d)). At the same time, the procedure resulted in increasing the total panel count by 45 for the 40-cm panels and by 74 for the 36-cm panels.

It was expected that the optimal panelization will be achieved with the smallest panel width. And as expected, the narrowest panels provided the closest fit in the base structure, although, they generated the biggest number of short cut-offs that could not be adapted elsewhere. Surprisingly, for the tested building the most efficient panelization in terms of waste minimization turned out to be the panelization with the 40 cm width standard panel. In the panelization with 40-cm panels, over 50% of cut-offs were equal or wider than 20 cm and could be reused. Accordingly, the number of manufactured unique panels could be decreased. However, the resultant increase of total number of panels could translate into longer construction time.

Discussion

The development of digital tools and techniques that came with the Fourth Industrial Revolution raised the quality of architects’ deliverables. Thanks to simulation and generative methods, architects received means to significantly improve the performance of their designs. In the presented scenario, the preliminary non-optimal panelization could be improved and instead of time-consuming rework, the process could have been automated. The process automation allowed for more precise division as well as for exploring multiple options. The presented results revealed that automating the panelization phase could contribute to reducing the amount of waste generated during production and assembly of the small-element timber house. This contribution could be made by optimal selection of prefabricated panels sizes. Considering that the proposed solution is scalable, the material efficiency could be even more distinct in larger developments.

In practice, the subdivision process is often performed manually. Thus, it is performed after concluding the design phase. Due to time limitations, the resulting panelization is usually nonoptimal, since adjusting it would involve repeating the whole procedure. The automated BIM-based simulation eliminated this issue and, therefore, allowed for improving the decision making in early design stages. The use of BIM environment enabled integrating all the design stages in a single model. The panelization process was based on data retrieved from the model. At the same time, the data achieved in the simulation were then used to alter the original model. The original BIM model could be utilized to compare different options based on the introduced optimization criteria. Moreover, the architect was able to design the building without worrying about the prefabrication constraints and could focus on functionality and aesthetics.

Implementing solid timber structure increased the flexibility potential of the building as well as its sustainability, since timber is a carbon-negative material and, due to moisture diffusion, provides a good living microclimate. In the context of global changes and uncertain times such as global warming and shrinking resources, turning towards renewable and climate-friendly materials is a crucial step in housing development. Simultaneously, it is equally essential to use the available resources responsibly and to avoid wasting them. This is a field where BIM proved to be a useful methodology. The data stored in the model facilitated simulating the actual need for the structural material and optimizing the panelization so that the generated waste was minimal.

The proposed solution was meant to aid adapting the design to the requirements of the DIY assembly by generating prefabricated elements in sizes suitable for manual handling. The DIY approach responds to the increasing demand for giving the end user freedom in terms of adjusting the building to their needs changing in time. By limiting the number of qualified workers in the construction process, the solution could decrease the total cost of investment, which addresses the urge for affordable housing. This approach converges with Peter Ward’s²² observation that self-help is a significant driver of low-income housing development, especially in the areas of rapid urbanization. To further enhance the DIY-potential, the small-element technology introduced in this paper could benefit from developing easier assembly methods such as snap-fit joints described by Kareem Elsayed et al.²³ Interlocking connectors could replace nails and screws and therefore further reduce the construction time. The connectors’ geometry could be represented in the model and then be used in the manufacturing phase.

Although the problem of manufacturing was not considered in this research, the implementation of BIM methodology could be utilized to link the design phase with digitally aided fabrication. For wood manufacturing, the commonly used data interface is Building Transfer Language (BTL).²⁴ A BTL file can contain a comprehensive information specific for automated fabrication of complex timber shapes, however, currently the BTL standard is not supported by most of the commonly used BIM platforms. In Autodesk Revit, which was utilized in the presented research, the BTL file can be exported via a specialized extension. Alternatively, a BIM model could be transferred to CAM software using Industry Foundation Classes (IFC) format. Although IFC does not cover the processes related to model elements, as an interoperability standard, it allows other geometric applications to exchange information with a BIM tool. An IFC file can handle models with high level of detail, which can be sufficient for fabrication.²⁴

Finally, the presented tool has the potential of complementing research in other areas of computationally aided architectural design. The simulation of prefabricated structure in real time could be combined with automated generation of house layouts in the participatory process as proposed by Krystian Kwieciński.²⁵ A comparable participatory approach was developed by Joseph Huang and Robert Krawczyk.²⁶ The authors demonstrated a web-based system for assisting the architect and the client in selection of the most suitable option of a prefabricated single family house design. The additional benefit of integrating the design within BIM could further enhance the decision making by benchmarking the realization cost and constriction time. Similar utilization of BIM was introduced by Verley Côco and Gabriela Celani,²⁷ who researched automated generation of prefabricated bathroom layouts combined with automatic documentation for manufacturing purposes.

Conclusion

The impact of the Fourth Industrial Revolution on architecture gradually grows. Incorporating advanced digital solutions in the design process becomes a standard in architectural practice. Automating repeatable tasks, along with developing custom computational solutions, conducting simulations and optimizing design outcomes serve for improving the performance of architectural realizations at all design scales. The aim of the presented research was to explore the potential of combining the advancements of the Fourth Industrial Revolution such as BIM and generative design with small scale projects, that is, single-family houses. Due to its complexity and multitude of features, BIM is commonly associated with large realizations that involve many disciplines, a complex coordination and require accurate data management. The presented case aimed to demonstrate that also a single-family house design could involve complex problems, such as optimal structure subdivision, which could benefit from using an integrated BIM approach. The introduced prototype utilized multiple BIM features: model geometry, quantitative data and creating documentation, and combined them in a generative procedure in order to speed up the design process and improve the overall quality of the outcome.

The growing number of investments involving timber prefabrication allows to suspect that this trend will continue. Timber prefabricates gain popularity in all scales of constructions: from single family houses to high-rise residential and office buildings. Even though timber itself contributes to reducing the environmental impact of the investment, there is always room for improvement. This improvement could be achieved by developing new dedicated tools and methods as well as integrating already existing research into more advanced solutions that support digital prototyping. As pointed out by Odysseas Kontovourkis,²⁸ the concept of sustainability in the field of architecture, engineering and construction (AEC) is inseparably tied to advanced tools and their effective implementation, as was the intent of the presented prototype. It adapts a standard BIM model for the purpose of timber prefabrication by extracting data necessary for simulating and optimizing the panelization. It enabled decreasing the amount of generated waste, which translated into improving the building’s sustainability.

Finally, the DIY approach in architecture is commonly associated with small-scale realizations such as single-family houses. The introduced tool contributed to this field by adjusting the prefabricated panel sizes to the limitations of manual assembly. Moreover, one can imagine building a house on his or her own, whereas in terms of multi-family housing or other building functions it is not commonly practised. However, at all scales one feature remains valuable: flexibility. Adjusting a living space to the needs changing over time is often a challenge. Introducing a DIY-oriented prefabrication might address this problem, since partitions of timber panels can be easily removed and installed elsewhere. The presented tool could be an asset while preparing an investment, supplemented with an instruction for possible DIY alterations.

Although this paper addresses strictly small-element timber prefabrication, the introduced tool is a part of wider research exploring the potential of BIM in design with the use of prefabrication. It is planned to further develop the prototype into a comprehensive aid for participatory design of single-family houses and, furthermore, also multi-family developments realized by housing cooperatives. The aim of future work is to create a multidisciplinary interface bringing together sustainability of design, architects’ creative freedom, structural and fabrication requirements and the end user’s needs.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Karolina Ostrowska-Wawryniuk

References

Smith

RE.

Prefab architecture. a guide to modular design and construction. Hoboken, NJ: Wiley, 2010.

Meuser

Construction and design manual. Prefabricated housing. Volume 2: buildings and typologies. Berlin: DOM Publishers, 2018.

Zarecor

Manufacturing a socialist modernity: housing in Czechoslovakia, 1945-1960. Pittsburgh: University of Pittsburgh Press, 2011.

Meuser

Zadorin

Towards a typology of Soviet mass housing. Prefabrication in the USSR 1955 – 1991. Berlin: DOM Publishers, 2016.

Schneider

Till

Flexible housing. London: Routledge, 2007.

McKean

Learning from Segal: Walter Segal’s life, work and influence. Boston: Birkhäuser, 1989.

Ostrowska-Wawryniuk

Stępniak

Modelowy dom drewniany. In: Giecewicz

(ed.) Mieszkania jako problem publiczny, społeczny, prywatny? Warszawa: Wydział Architektury Politechniki Warszawskiej, 2018, pp.23–43.

Salingaros

Tejada

DM.

Modularity and the number of design choices. Nexus Netw J 2001; 3(1): 99–109.

Bock

Linner

Robotic industrialization: automation and robotic technologies for customized component, module, and building prefabrication. Cambridge: Cambridge University Press, 2015.

10.

Bermek

Shelden

Gentry

TR.

A holistic approach to feature-based structural mapping in cross laminated timber buildings. In: Architecture in the age of the 4th industrial revolution - proceedings of the 37th eCAADe and 23rd SIGraDi conference, 2019, pp.789–796. Porto: eCAADe, SIGraDi, Faculty of Architecture, University Porto.

11.

Neumann

Schmidt

Mayor

. CNC timber framing – innovative applications of digital wood fabrication technology. In: SIGraDi 2006, 2006, pp.304–307.

12.

Cormack

Sweet

. Parametrically fabricated joints: creating a digital workflow. In: SIGraDi 2016, 2016, pp.412–417. Buenos Aires: SIGraDi.

13.

Chai

Marino

, et al. Design for mass customization robotic realization of a timber tower with interlocking joints. In: Acadia 19: ubiquity and autonomy, 2019, pp.564–572. Austin: ACADIA.

14.

Patlakas

Livingstone

Hairstans

. A BIM platform for offsite timber construction. In: Real time: extending the reach of computation - proceedings of the 33rd eCAADe conference, 2015, pp.597–604. Vienna: eCAADe.

15.

Mekawy

Petzold

. Exhaustive exploration of modular design options to inform decision making. In: ShoCK! - sharing computational knowledge! - proceedings of the 35th eCAADe conference, 2017, pp.107–114. Rome: eCAADe.

16.

Mekawy

Petzold

. BIM-based model checking in the early design phases of precast concrete structures. In: 23rd International conference on computer-aided architectural design research in Asia: learning, prototyping and adapting, 2018, pp.71–80. Beijing: CAADRIA.

17.

Ostrowska-Wawryniuk

Nazar

. Generative BIM automation strategies for prefabricated multi-family housing design. In: Computing for a better tomorrow - proceedings of the 36th eCAADe conference, 2018, pp.247–256. Łódź: Lodz University of Technology.

18.

Martins

De Campos

Nunes

, et al. Expanding the material possibilities of lightweight prefabrication in concrete through robotic hot-wire cutting - form, texture and composition. In: Real time: extending the reach of computation - proceedings of the 33rd eCAADe conference, 2015, pp.341–351. Vienna: eCAADe.

19.

Meibodi

Jipa

Giesecke

, et al. Smart slab: Computational design and digital fabrication of a lightweight concrete slab. In: Recalibration on imprecision and infidelity - proceedings of the 38th annual conference of the association for computer aided design in architecture, 2018, pp.434–443.

20.

Hüttemann. Glulam timber elements. Wismar: Hüttemann Holz GmbH & Co. KG, 2007, pp.1–30.

21.

Ostrowska-Wawryniuk

. BIM-aided prefabrication for minimum waste DIY timber houses. In: Architecture in the age of the 4th industrial revolution - proceedings of the 37th eCAADe and 23rd SIGraDi conference, 2019. pp.251–258. Porto: eCAADe, SIGraDi, Faculty of Architecture, University of Porto.

22.

Ward

. Self-help housing. In: The wiley blackwell encyclopedia of urban and regional studies. Hoboken: Wiley, 2019, pp.1–6.

23.

Elsayed

Fioravanti

Squasi

Low-cost housing: testing snap-fit joints in agricultural residue panels. In: ShoCK! - sharing computational knowledge! - proceedings of the 35th eCAADe conference, vol. 2, 2017, pp.167–174.

24.

Staub-French

Poirier

, et al. Building information modeling (BIM) and design for manufacturing and assembly (DfMA) for mass timber construction. Vancouver: BIM TOPiCS Research Lab, University of British Columbia, 2018.

25.

Kwiecinski

Slyk

. System for customer participation in the design process of mass-customized houses. In: Fusion: data integration at its best - proceedings of the 32nd eCAADe conference, vol 2, 2014, pp.207–215. Newcastle: eCAADe.

26.

Huang

Krawczyk

i_Prefab home - customizing prefabricated houses by internet-aided design. In: Commun space(s) proceedings of the 24th eCAADe conference, 2006, pp.690–698. Volos: Volos University of Thessaly.

27.

Côco Júnior

Celani

. From the automated generation of layouts to fabrication with the use of BIM: a new agenda for architecture in the 21st century. In: SIGraDi 2018, 2018, pp.23–30. São Paulo: SIGraDi.

28.

Kontovourkis

Multi-objective design optimization and robotic fabrication towards sustainable construction. In: ShoCK! - sharing computational knowledge! - proceedings of the 35th eCAADe conference, 2017, pp.125–134. Rome: eCAADe.