495 Supporting Tools for Early Stages of Architectural Design Ernesto Bueno and Benamy Turkienicz issue 4, volume 12 international journal of architectural computing
Supporting Tools for Early Stages of Architectural Design
Mar 16, 2023
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
00_Contents495
Supporting Tools for Early Stages of Architectural Design Ernesto Bueno and Benamy Turkienicz
issue 4, volume 12international journal of architectural computing
496
Supporting Tools for Early Stages of Architectural Design Ernesto Bueno and Benamy Turkienicz
In architectural design, pencil and paper remain the most frequently used media to create freehand drawings to support early design stages. Digital tools conventionally used by architects lack appropriate functionalities and do not offer friendly interfaces for the early stages of architectural design.These are the bad news.The good news are twofold: a) hardware already available can help freehand designers to digitally express their first ideas; and b) functionality principles present in experimental software combined with appropriate hardware could successfully provide a friendly and intuitive human-computer interaction in the early stages of architectural design.This paper takes special attention to the way architects interact with computers, how input devices constrain possible interactions and how functionalities can be explored through these interactions.The article summarizes basic principles to be considered in the development of an all-in-one software and create a scenario whereby these principles are simulated on a hypothetical software to be used during the early stages of architectural design.
1. INTRODUCTION
As a design activity, drawing is a visual and mental transaction, a conversation that the designer establishes with him/herself.The ideation process is shaped by the reflexive cycle: sketch, inspect, revise [1]. Design improvements are achieved from the designer’s sequence of actions resulting in transformations of the representation [2].To make progress through the reflexive cycle, designers must have the ability to transform implicit knowledge into representational structures fast enough, so that new changes and amendments reflect the interpretation of what was revised, creating continuity between mind and the drawing in process [3].
To establish this connection between mind and drawing, designers make use of their visual intelligence [4] as to associate different represented shapes [5].Visual intelligence is one of the main types of human intelligence connected with major parts of the brain [6]. It is specially developed in architects, who exploit this potential through visual training during their education and professional experience [4].The continuous training allows architects to depict emergent shapes and patterns which are fundamental for the intermittent reinterpretation of a design idea. In architectural design, visual emergence set the environment for shape interpretations of architectural elements. For example, in Figure 1, the designer recognizes, in the overlapping of two rectangles, an emergent rectangular shape.This can be interpreted as a shadowed area at the overlapping of rectangular volumes (Figure 1, Left), or as a void, framed by two L-shaped volumes (Figure 1, Right).These emergent shapes can be interpreted as architectural elements with different spatial properties existing in a (not yet represented) third dimension, as illustrated in Figure 2.The first interpretation would define an internal space without direct sunlight (Figure 3, Left), while the second interpretation would define internal spaces with direct sunlight (Figure 3, Right).
Figure 1: Manual recognition of
emergent shapes on conceptual
illustrating architectural consequences
of different interpretations.
497Supporting Tools for Early Stages of Architectural Design
The sequence of drawings acts as a registry of the process of reasoning that designers build from inspection and revision of previous drawings, which demarcates the thresholds from one representation to another.The “re- representation”, or sequence of representations produced through over sketching, can result in proposals of unexpected novelty and diversity [2]. By inspecting the sequence of drawings, designers can reinterpret relations and features which were not initially clear, suggesting ways of revising original ideas: through re-representation, the new is built upon the existent [2].
While focusing on thinking and representing design solutions within the reflexive cycle, designers are not used to explicate the knowledge and abilities applied to design, nor how these knowledge and abilities are used [7].Architectural designers implicitly coordinate a series of personal methods based on the experience of constructive perception [8], an ability of the mind that allows trained designers to coordinate cognitive actions of reorganization (through visual emergence) and conceptual generation (by re-representation). Incorporating this specific ability, designers improve personal methods whereby they achieve design solutions and, through experience, get faster and better connected with their intuitive world [8].
To shed light over some basic requirements that digital technologies could offer to architects at early design stages, this article is divided in six sections. Section 2, describe existing representational methodologies in the early stage design. Section 3 demonstrates how designers interact with software and hardware technologies in architectural design and some of the difficulties architects suffer in dealing with current tools and techniques. In section 4, the state-of-the-art studies of experimental technologies supporting interactions and functionalities are reviewed. Section 5 describe a scenario of use of a hypothetical software for early stages of architectural design. In section 6, the basic principles for development of such an application are briefly suggested.
2. DESIGN REPRESENTATIONS IN THE ARCHITECTURAL DESIGN PROCESS
The most representative workflow of an architect begins with contextual data collection, formulation of the design problem and developing sequences of sketches that progressively address formulated problems [2]. In the early stages, design problems are so ambiguous that freehand drawings are
Figure 3: Perspectives showing 3D
models of different sketch
498 Ernesto Bueno and Benamy Turkienicz
preferred [9] when designers take advantage of their ability to reorganize emerging shapes [10], coordinating them in conceptual generation [8] of re- representations [2], and defining the architectural team. In this phase, sketch remains the most appropriate representation to address design problems [11, 12, 13], followed by more accurate technical drawings, using CAD [11].
Activities in the early architectural design stages typically happen through sketches using pencil and paper [11, 14]. Ideas shaped in paper externalize, through fuzzy or vague representations, the designer’s intentions in the approach of ill-defined problems. In advanced stages, designs are digitally represented with technical drawings [15], using CAD tools, mainly Autodesk AutoCAD; 3D models, typically using Trimble SkekcthUp [16]; and performance simulations, using software such as Autodesk Ecotect [17]. Performance simulation software includes resources that temporarily change the appearance of the model in different ways, according to the simulated phenomena: airflow is represented by drawing vector fields; lighting simulations involve the discretization of models by sampling and coloring polygon meshes, according to the incident light. CAD, 3D modeling and simulation software, offer features targeted for more accurate representations than those produced with freehand drawings.To use the digital media, it is necessary to transfer architectural design ideas sketched on paper to computers and specify the information, making explicit issues that, in sketches, were diffusely or implicitly addressed [18].
Designers transfer design ideas from paper to digital environments by scanning and redrawing in CAD, overlapping layers above bitmap images of the sketches; or creating new CAD drawings from scratch, based on observation and re-creation of features from the sketch [19].The transfer process involve interpretation, is time consuming [11] and inefficient [13].
The inefficiency of media transfer can be avoided, by eliminating the need to transfer. In other words, if all stages are developed in the digital environment [11] designers will employ freehand drawing techniques for the creation of early design sketches.The problem is that existing input devices such as styli [20] and the bitmap images generated with these hardware and aggregated software are not compatible with the requirements of vector drawings in CAD.
There exists a trend, in architectural design, for using CAD tools in all design stages [11, 14].Aside from the underuse of styli in CAD, there are several limitations in CAD tools in comparison to freehand drawing.This trend is detrimental to the architectural project for reasons which will be explained as follows.
2.1. Drawbacks in the use of CAD tools in early stages of design
1. Additional cognitive load makes the design process inefficient: the connection between mind and CAD drawing through the mouse makes the initial design activities slow and inefficient, compared to
499Supporting Tools for Early Stages of Architectural Design
the connection between mind and sketch through the pencil [13]. The pencil allows a continuous drawing process, even on ill-defined ideas.When stopping the process to follow CAD command protocols, designers increase their cognitive load [21] and take longer to define a design solution [11, 22].
2. Communication gets disturbed: when the designer works without prior creation of hand drawings, the collaboration between designers and communication with the client (or teacher) are difficult, since the visual discourse based on graphic conventions is broken [14]. Conventions, such as the visual gestures of communicative sketches and architectural symbols present in prescriptive sketches, enrich communications of design ideas and avoid misinterpretations [23].
3. The quality of the design diminishes: projects initiated in CAD are often evaluated as worse than projects initiated with freehand drawings, as registered by instructors of Architectural Design Studios [11, 14, 24].
Summing up, the use of CAD tools in the early design stages turn design into a slow and inefficient process, with design solutions lacking communicative qualities and worse than those initiated with freehand sketches [11, 14, 24].To support the continuity between design stages it is necessary that, in the early stages, representations are created with digital tools supporting freehand sketching, similar to paper and pencil [25]. Conceptual ideas will be represented faster, closer to what the designer imagine and will offer a more adequate communication flow than current CAD and 3D modeling tools [25].
3. INTERACTION WITH INPUT DEVICES AND PERFORMANCE SIMULATIONS IN ARCHITECTURAL DESIGN
Technologies supporting architectural design do not widely offer (to the end-user) digital tools for the most important activities in the early stages of the design process, especially freehand drawing [25]. Interaction with desktop and laptop computers remain, almost exclusively, based on mouse and keyboard input.This interaction suffer with limitations, starting with the design of the mouse.
3.1. Mouse interaction
Affordance is a property in the interaction between the user and the device, based on the perception of the characteristics of the device [21].The mouse’s affordance is related to pointing and clicking buttons, as required for CAD drawing [26]. Examining the hand’s position above the mouse and the possible movements of the mouse in relation to the movements perceived by the user on the screen, it is possible to affirm that this device
500 Ernesto Bueno and Benamy Turkienicz
does not offer a good relationship of perceived control of free pointer drag. Drag is the action required to create the strokes that form the sketch [27]. From its drag function’s point of view, the mouse is inappropriate for drawing sketches [28].
The mouse, located outside the visual field in which the relation cognition-representation is established, oblige users to make an extra effort of synchronization between the movement of the hand and the corresponding movement of the pointer on the screen, as two distinct agents, responding to different rules, acting in the same activity [26].
In the cycle: draw, inspect, revise, for each design iteration, two physical gestures are, at least, required: the user moves the mouse, inspects the consequent movement of the pointer, providing a feedback of the first movement, which informs motor adjustments in a second mouse movement, with which the edition is started.This interposition limits continuity between cognition and representation, which is a key factor for the development of the design through iterations [29].
3.2. Pen and multi-touch interaction
Company et al. [13] demonstrated the qualities of interaction through stylus pen, with a comparative study between the aforementioned tools and traditional media.The study focused on the assessment of efficiency, according to the time spent on sketching, and usability, from the qualitative response from users. Sketching with pencil and paper was judged as more intuitive and faster than how CAD drawing was, especially between experienced designers. It was noted that pencils allow improvisations and imperfections and that are comfortable to the hand.The same work done with stylus on a tablet computer, was reported as advantageous for its similarity to traditional pencils, sharing the same affordance for writing and sketching, in addition of being fast, clean and easy to erase [13].
The adoption of tablet computers, in addition to promoting the use of styli, contributes to the familiarization of multi-touch gestures, generally used in text and image applications, but with principles of use common to associated design tasks, especially navigation (e.g., zoom and pan) and editing (e.g., moving, rotating, scaling).
Hinckley et al. [30, 31], address the problem of developing a software for graphical annotations (diagrams and storyboards) that, exploring these technologies, offer an alternative bimanual interaction by combining pen and multi-touch simultaneously.This blend was inspired in the creative activities of bimanual work made with traditional media, combining tools such as pencils, rulers, scissors and manipulation of sheets of paper.The study has shown that, for users, pen and multi-touch are perceived as easy to use as traditional tools and media [31]. Considering the benefits of working in digital environments, such as data storage and portability, communication features, latter reproduction and edition and, most importantly, the
501Supporting Tools for Early Stages of Architectural Design
possibility of direct transition between the early and advanced stages, supporting the appropriate CAD formats, Hinckley’s conclusions seem to constitute an important support for the use of digital media in the early stages of architectural design.
3.3. Model affordance from performance simulations
In the decision making process at different design stages, the designer interacts with the representation in the construction of the design idea whereby the designer applies implicit knowledge to improve the design [2]. The designer´s knowledge related to the building performance under development is externalized through architectural drawings [32]. By using performance simulation software, the designer utilizes his/her visual intelligence to relate implicit knowledge with performance values represented in the simulation as to make informed decisions [33].When performance results are unfavorable, rethinking the design is required.The visualization of the impact of the simulated characteristics (e.g., the dimensions of the massing), help the designer to reshape the architectural form which, in turn, will update the simulation results [34].The relationship between simulation and design can be seen as the cyclical interaction between the designer and the simulation model, based on the perception of the characteristics simulated [33].The performance simulations emulate the architectural model affordances towards design improvements [21].As architectural affordances are better explored during the early stages of design, initial design decisions are likely to affect the building performance more intensely than at the intermediate and final design stages. Modifications to improve building performance, involve more specific and costly data in the intermediate and final design stages than in the early stages and it is more time-consuming.
3.4.Availability of performance simulation, pen and multi-touch technologies for architects
Architects and the building industry will always be interested to perform simulations in the early stages of design, when design modifications are not as expensive as the intermediate and final design stages [35]. For this purpose,Autodesk introduced Ecotect for the analysis of the buildings environmental performance [36] and proposed Vasari to assist designers in the early stages of architectural design [17]. Paradoxically,Vasari was designed to work jointly with software to support advanced stages of design and construction (i.e., BIM). In order to grant its BIM compatibility, Vasari requires the addressing of specific building attributes [17].This automatically drives the designer attention away from the conceptual problems, breaking the mental workflow of the early design stages.
Hardware for direct input such as multi-touch screens and styli, allowing digital freehand drawings, are commercially available for the general end-
502 Ernesto Bueno and Benamy Turkienicz
user and widely used since the introduction of smartphones and tablet computers [37, 38, 41].
With the exception of Autodesk FormIt [42], available applications with pen and multi-touch interaction targeted to architects do not support multi-touch 3D modeling [43] and do not address an automatic transition from the early to the advanced stages of architectural design.With specific software out of the reach of architects, direct input devices are not seen nor experienced as design tools.There exists however, some experimental technology that can be used in the early design stages that have the potential to fulfil this gap as will be examined in the next section.
4. STATE-OF-THE-ART DIGITAL TOOLS FOR THE EARLY STAGES OF ARCHITECTURAL DESIGN
The state of the art on digital tools which could potentially be used for the early stages of architectural design had been reviewed, classified and schematically synthetized in Table 1. In this table, the main features of experimental software are organized and compared.All compared digital
Table 1: Comparison of state-of-the-
art experimental software main
Reference Application Area
Automatic N/A High (DXF)
2D vector drawing
N/A N/A N/A
Juchmes et al., 2005 [25] Architecture Pen 2D sketching Automatic N/A N/A
Kallio, 2005 [29] Architecture & Design
Pen & keyboard
Masry & Lipson, 2005 [19] Engineering Pen Perspective recognition
Automatic Structural N/A
3D sketch extrusion
WIMP N/A N/A
3D sketching N/A N/A Low (2D)
Elsen & Leclerq, 2008 [48] Architecture Pen 2D sketching N/A N/A N/A
Naya et al., 2008 [49] Engineering Pen 3D sketching Gestural N/A High (DXF)
Lopes et al., 2011 [50] Visual Arts Pen & multi- touch
3D sketching Gestural N/A High (3D)
Elsen et al., 2012 [51] Architecture Pen 2D sketching Automatic N/A N/A
Kang et al., 2013 [43] Engineering Multi-touch 3D sketching Gestural N/A Medium (macro)
tools feature sketch acquisition and, with the exception of Kallio [29] and Dorsey et al. [44], all perform some kind of shape recognition.
It can be observed that two digital tools [39, 47] base the user’s interaction on mouse and keyboard.Yu and Zhang [18] define the user’s interaction only with a pointer, for latter replace of the mouse with a pen. Most of tools define drawing interaction with the pen, but Kallio [29] and Dorsey et al. [44] rely on the keyboard to control the 3D positioning. Kang et al. [43] support the design solely with the fingers with multi-touch gestures, Lopes et al. [50] explore a multimodal, bimanual interaction [30, 31] combining pen and multi-touch, and Pranovich [46] adapts 2D vector drawing for the development of schematic architectural plans. Company et al. [45], and Masry and Lipson [19] implement a complex algorithm to recognize 3D shapes automatically from perspective sketches, while Juchmes et al. [25] and Elsen et al. [51] address the automatic reconstruction from floor plan sketches.
Kallio [29] and Dorsey et al. [44] do not use solid geometry but 3D sketches, avoiding the implementation of reconstruction routines.The construction of solid geometry, proposed by Oh et al. [47], and Yu and Zhang [18], involve the extrusion of recognized polygons by interacting with WIMP interface resources (mainly buttons). Naya et al. [49], Lopes et al. [50] and Kang et al. [43] propose supporting gestures for the sketch reconstruction control. Except for Masry and Lipson [19], preliminary performance simulations of building design are not explored.
Company et al. [45] and Naya et al. [49] offer high interoperability by using the DXF format, whereas Lopes et al. [50] offer high interoperability with other 3D formats. Kallio [29] provides interoperability through VRML files with generated data limited to 3D lines. Oh et al. [47] permit external performance simulation through exported models, while Dorsey et al. [44] allow the generation of perspective images, which cannot be used in subsequent design stages.
All the functionality principles mentioned should be integrated in one single software for an appropriated support to early stages of architectural design. Each of these functionality principles have a specific role or attribute needed in software for the early stages of architectural design and are synthetically presented in Figure 4 and described as follows.
Figure 4: Principles of functionality
needed in software for early stages of
architectural design.
4.1. Functionality principles
Multimodal, bimanual interaction: Interaction modes, inspired in traditional bimanual work—with stylus pen and multi-touch screens—define three interaction modes: (i) Creation: drawing and annotating with a stylus [49]; (ii)…
Supporting Tools for Early Stages of Architectural Design Ernesto Bueno and Benamy Turkienicz
issue 4, volume 12international journal of architectural computing
496
Supporting Tools for Early Stages of Architectural Design Ernesto Bueno and Benamy Turkienicz
In architectural design, pencil and paper remain the most frequently used media to create freehand drawings to support early design stages. Digital tools conventionally used by architects lack appropriate functionalities and do not offer friendly interfaces for the early stages of architectural design.These are the bad news.The good news are twofold: a) hardware already available can help freehand designers to digitally express their first ideas; and b) functionality principles present in experimental software combined with appropriate hardware could successfully provide a friendly and intuitive human-computer interaction in the early stages of architectural design.This paper takes special attention to the way architects interact with computers, how input devices constrain possible interactions and how functionalities can be explored through these interactions.The article summarizes basic principles to be considered in the development of an all-in-one software and create a scenario whereby these principles are simulated on a hypothetical software to be used during the early stages of architectural design.
1. INTRODUCTION
As a design activity, drawing is a visual and mental transaction, a conversation that the designer establishes with him/herself.The ideation process is shaped by the reflexive cycle: sketch, inspect, revise [1]. Design improvements are achieved from the designer’s sequence of actions resulting in transformations of the representation [2].To make progress through the reflexive cycle, designers must have the ability to transform implicit knowledge into representational structures fast enough, so that new changes and amendments reflect the interpretation of what was revised, creating continuity between mind and the drawing in process [3].
To establish this connection between mind and drawing, designers make use of their visual intelligence [4] as to associate different represented shapes [5].Visual intelligence is one of the main types of human intelligence connected with major parts of the brain [6]. It is specially developed in architects, who exploit this potential through visual training during their education and professional experience [4].The continuous training allows architects to depict emergent shapes and patterns which are fundamental for the intermittent reinterpretation of a design idea. In architectural design, visual emergence set the environment for shape interpretations of architectural elements. For example, in Figure 1, the designer recognizes, in the overlapping of two rectangles, an emergent rectangular shape.This can be interpreted as a shadowed area at the overlapping of rectangular volumes (Figure 1, Left), or as a void, framed by two L-shaped volumes (Figure 1, Right).These emergent shapes can be interpreted as architectural elements with different spatial properties existing in a (not yet represented) third dimension, as illustrated in Figure 2.The first interpretation would define an internal space without direct sunlight (Figure 3, Left), while the second interpretation would define internal spaces with direct sunlight (Figure 3, Right).
Figure 1: Manual recognition of
emergent shapes on conceptual
illustrating architectural consequences
of different interpretations.
497Supporting Tools for Early Stages of Architectural Design
The sequence of drawings acts as a registry of the process of reasoning that designers build from inspection and revision of previous drawings, which demarcates the thresholds from one representation to another.The “re- representation”, or sequence of representations produced through over sketching, can result in proposals of unexpected novelty and diversity [2]. By inspecting the sequence of drawings, designers can reinterpret relations and features which were not initially clear, suggesting ways of revising original ideas: through re-representation, the new is built upon the existent [2].
While focusing on thinking and representing design solutions within the reflexive cycle, designers are not used to explicate the knowledge and abilities applied to design, nor how these knowledge and abilities are used [7].Architectural designers implicitly coordinate a series of personal methods based on the experience of constructive perception [8], an ability of the mind that allows trained designers to coordinate cognitive actions of reorganization (through visual emergence) and conceptual generation (by re-representation). Incorporating this specific ability, designers improve personal methods whereby they achieve design solutions and, through experience, get faster and better connected with their intuitive world [8].
To shed light over some basic requirements that digital technologies could offer to architects at early design stages, this article is divided in six sections. Section 2, describe existing representational methodologies in the early stage design. Section 3 demonstrates how designers interact with software and hardware technologies in architectural design and some of the difficulties architects suffer in dealing with current tools and techniques. In section 4, the state-of-the-art studies of experimental technologies supporting interactions and functionalities are reviewed. Section 5 describe a scenario of use of a hypothetical software for early stages of architectural design. In section 6, the basic principles for development of such an application are briefly suggested.
2. DESIGN REPRESENTATIONS IN THE ARCHITECTURAL DESIGN PROCESS
The most representative workflow of an architect begins with contextual data collection, formulation of the design problem and developing sequences of sketches that progressively address formulated problems [2]. In the early stages, design problems are so ambiguous that freehand drawings are
Figure 3: Perspectives showing 3D
models of different sketch
498 Ernesto Bueno and Benamy Turkienicz
preferred [9] when designers take advantage of their ability to reorganize emerging shapes [10], coordinating them in conceptual generation [8] of re- representations [2], and defining the architectural team. In this phase, sketch remains the most appropriate representation to address design problems [11, 12, 13], followed by more accurate technical drawings, using CAD [11].
Activities in the early architectural design stages typically happen through sketches using pencil and paper [11, 14]. Ideas shaped in paper externalize, through fuzzy or vague representations, the designer’s intentions in the approach of ill-defined problems. In advanced stages, designs are digitally represented with technical drawings [15], using CAD tools, mainly Autodesk AutoCAD; 3D models, typically using Trimble SkekcthUp [16]; and performance simulations, using software such as Autodesk Ecotect [17]. Performance simulation software includes resources that temporarily change the appearance of the model in different ways, according to the simulated phenomena: airflow is represented by drawing vector fields; lighting simulations involve the discretization of models by sampling and coloring polygon meshes, according to the incident light. CAD, 3D modeling and simulation software, offer features targeted for more accurate representations than those produced with freehand drawings.To use the digital media, it is necessary to transfer architectural design ideas sketched on paper to computers and specify the information, making explicit issues that, in sketches, were diffusely or implicitly addressed [18].
Designers transfer design ideas from paper to digital environments by scanning and redrawing in CAD, overlapping layers above bitmap images of the sketches; or creating new CAD drawings from scratch, based on observation and re-creation of features from the sketch [19].The transfer process involve interpretation, is time consuming [11] and inefficient [13].
The inefficiency of media transfer can be avoided, by eliminating the need to transfer. In other words, if all stages are developed in the digital environment [11] designers will employ freehand drawing techniques for the creation of early design sketches.The problem is that existing input devices such as styli [20] and the bitmap images generated with these hardware and aggregated software are not compatible with the requirements of vector drawings in CAD.
There exists a trend, in architectural design, for using CAD tools in all design stages [11, 14].Aside from the underuse of styli in CAD, there are several limitations in CAD tools in comparison to freehand drawing.This trend is detrimental to the architectural project for reasons which will be explained as follows.
2.1. Drawbacks in the use of CAD tools in early stages of design
1. Additional cognitive load makes the design process inefficient: the connection between mind and CAD drawing through the mouse makes the initial design activities slow and inefficient, compared to
499Supporting Tools for Early Stages of Architectural Design
the connection between mind and sketch through the pencil [13]. The pencil allows a continuous drawing process, even on ill-defined ideas.When stopping the process to follow CAD command protocols, designers increase their cognitive load [21] and take longer to define a design solution [11, 22].
2. Communication gets disturbed: when the designer works without prior creation of hand drawings, the collaboration between designers and communication with the client (or teacher) are difficult, since the visual discourse based on graphic conventions is broken [14]. Conventions, such as the visual gestures of communicative sketches and architectural symbols present in prescriptive sketches, enrich communications of design ideas and avoid misinterpretations [23].
3. The quality of the design diminishes: projects initiated in CAD are often evaluated as worse than projects initiated with freehand drawings, as registered by instructors of Architectural Design Studios [11, 14, 24].
Summing up, the use of CAD tools in the early design stages turn design into a slow and inefficient process, with design solutions lacking communicative qualities and worse than those initiated with freehand sketches [11, 14, 24].To support the continuity between design stages it is necessary that, in the early stages, representations are created with digital tools supporting freehand sketching, similar to paper and pencil [25]. Conceptual ideas will be represented faster, closer to what the designer imagine and will offer a more adequate communication flow than current CAD and 3D modeling tools [25].
3. INTERACTION WITH INPUT DEVICES AND PERFORMANCE SIMULATIONS IN ARCHITECTURAL DESIGN
Technologies supporting architectural design do not widely offer (to the end-user) digital tools for the most important activities in the early stages of the design process, especially freehand drawing [25]. Interaction with desktop and laptop computers remain, almost exclusively, based on mouse and keyboard input.This interaction suffer with limitations, starting with the design of the mouse.
3.1. Mouse interaction
Affordance is a property in the interaction between the user and the device, based on the perception of the characteristics of the device [21].The mouse’s affordance is related to pointing and clicking buttons, as required for CAD drawing [26]. Examining the hand’s position above the mouse and the possible movements of the mouse in relation to the movements perceived by the user on the screen, it is possible to affirm that this device
500 Ernesto Bueno and Benamy Turkienicz
does not offer a good relationship of perceived control of free pointer drag. Drag is the action required to create the strokes that form the sketch [27]. From its drag function’s point of view, the mouse is inappropriate for drawing sketches [28].
The mouse, located outside the visual field in which the relation cognition-representation is established, oblige users to make an extra effort of synchronization between the movement of the hand and the corresponding movement of the pointer on the screen, as two distinct agents, responding to different rules, acting in the same activity [26].
In the cycle: draw, inspect, revise, for each design iteration, two physical gestures are, at least, required: the user moves the mouse, inspects the consequent movement of the pointer, providing a feedback of the first movement, which informs motor adjustments in a second mouse movement, with which the edition is started.This interposition limits continuity between cognition and representation, which is a key factor for the development of the design through iterations [29].
3.2. Pen and multi-touch interaction
Company et al. [13] demonstrated the qualities of interaction through stylus pen, with a comparative study between the aforementioned tools and traditional media.The study focused on the assessment of efficiency, according to the time spent on sketching, and usability, from the qualitative response from users. Sketching with pencil and paper was judged as more intuitive and faster than how CAD drawing was, especially between experienced designers. It was noted that pencils allow improvisations and imperfections and that are comfortable to the hand.The same work done with stylus on a tablet computer, was reported as advantageous for its similarity to traditional pencils, sharing the same affordance for writing and sketching, in addition of being fast, clean and easy to erase [13].
The adoption of tablet computers, in addition to promoting the use of styli, contributes to the familiarization of multi-touch gestures, generally used in text and image applications, but with principles of use common to associated design tasks, especially navigation (e.g., zoom and pan) and editing (e.g., moving, rotating, scaling).
Hinckley et al. [30, 31], address the problem of developing a software for graphical annotations (diagrams and storyboards) that, exploring these technologies, offer an alternative bimanual interaction by combining pen and multi-touch simultaneously.This blend was inspired in the creative activities of bimanual work made with traditional media, combining tools such as pencils, rulers, scissors and manipulation of sheets of paper.The study has shown that, for users, pen and multi-touch are perceived as easy to use as traditional tools and media [31]. Considering the benefits of working in digital environments, such as data storage and portability, communication features, latter reproduction and edition and, most importantly, the
501Supporting Tools for Early Stages of Architectural Design
possibility of direct transition between the early and advanced stages, supporting the appropriate CAD formats, Hinckley’s conclusions seem to constitute an important support for the use of digital media in the early stages of architectural design.
3.3. Model affordance from performance simulations
In the decision making process at different design stages, the designer interacts with the representation in the construction of the design idea whereby the designer applies implicit knowledge to improve the design [2]. The designer´s knowledge related to the building performance under development is externalized through architectural drawings [32]. By using performance simulation software, the designer utilizes his/her visual intelligence to relate implicit knowledge with performance values represented in the simulation as to make informed decisions [33].When performance results are unfavorable, rethinking the design is required.The visualization of the impact of the simulated characteristics (e.g., the dimensions of the massing), help the designer to reshape the architectural form which, in turn, will update the simulation results [34].The relationship between simulation and design can be seen as the cyclical interaction between the designer and the simulation model, based on the perception of the characteristics simulated [33].The performance simulations emulate the architectural model affordances towards design improvements [21].As architectural affordances are better explored during the early stages of design, initial design decisions are likely to affect the building performance more intensely than at the intermediate and final design stages. Modifications to improve building performance, involve more specific and costly data in the intermediate and final design stages than in the early stages and it is more time-consuming.
3.4.Availability of performance simulation, pen and multi-touch technologies for architects
Architects and the building industry will always be interested to perform simulations in the early stages of design, when design modifications are not as expensive as the intermediate and final design stages [35]. For this purpose,Autodesk introduced Ecotect for the analysis of the buildings environmental performance [36] and proposed Vasari to assist designers in the early stages of architectural design [17]. Paradoxically,Vasari was designed to work jointly with software to support advanced stages of design and construction (i.e., BIM). In order to grant its BIM compatibility, Vasari requires the addressing of specific building attributes [17].This automatically drives the designer attention away from the conceptual problems, breaking the mental workflow of the early design stages.
Hardware for direct input such as multi-touch screens and styli, allowing digital freehand drawings, are commercially available for the general end-
502 Ernesto Bueno and Benamy Turkienicz
user and widely used since the introduction of smartphones and tablet computers [37, 38, 41].
With the exception of Autodesk FormIt [42], available applications with pen and multi-touch interaction targeted to architects do not support multi-touch 3D modeling [43] and do not address an automatic transition from the early to the advanced stages of architectural design.With specific software out of the reach of architects, direct input devices are not seen nor experienced as design tools.There exists however, some experimental technology that can be used in the early design stages that have the potential to fulfil this gap as will be examined in the next section.
4. STATE-OF-THE-ART DIGITAL TOOLS FOR THE EARLY STAGES OF ARCHITECTURAL DESIGN
The state of the art on digital tools which could potentially be used for the early stages of architectural design had been reviewed, classified and schematically synthetized in Table 1. In this table, the main features of experimental software are organized and compared.All compared digital
Table 1: Comparison of state-of-the-
art experimental software main
Reference Application Area
Automatic N/A High (DXF)
2D vector drawing
N/A N/A N/A
Juchmes et al., 2005 [25] Architecture Pen 2D sketching Automatic N/A N/A
Kallio, 2005 [29] Architecture & Design
Pen & keyboard
Masry & Lipson, 2005 [19] Engineering Pen Perspective recognition
Automatic Structural N/A
3D sketch extrusion
WIMP N/A N/A
3D sketching N/A N/A Low (2D)
Elsen & Leclerq, 2008 [48] Architecture Pen 2D sketching N/A N/A N/A
Naya et al., 2008 [49] Engineering Pen 3D sketching Gestural N/A High (DXF)
Lopes et al., 2011 [50] Visual Arts Pen & multi- touch
3D sketching Gestural N/A High (3D)
Elsen et al., 2012 [51] Architecture Pen 2D sketching Automatic N/A N/A
Kang et al., 2013 [43] Engineering Multi-touch 3D sketching Gestural N/A Medium (macro)
tools feature sketch acquisition and, with the exception of Kallio [29] and Dorsey et al. [44], all perform some kind of shape recognition.
It can be observed that two digital tools [39, 47] base the user’s interaction on mouse and keyboard.Yu and Zhang [18] define the user’s interaction only with a pointer, for latter replace of the mouse with a pen. Most of tools define drawing interaction with the pen, but Kallio [29] and Dorsey et al. [44] rely on the keyboard to control the 3D positioning. Kang et al. [43] support the design solely with the fingers with multi-touch gestures, Lopes et al. [50] explore a multimodal, bimanual interaction [30, 31] combining pen and multi-touch, and Pranovich [46] adapts 2D vector drawing for the development of schematic architectural plans. Company et al. [45], and Masry and Lipson [19] implement a complex algorithm to recognize 3D shapes automatically from perspective sketches, while Juchmes et al. [25] and Elsen et al. [51] address the automatic reconstruction from floor plan sketches.
Kallio [29] and Dorsey et al. [44] do not use solid geometry but 3D sketches, avoiding the implementation of reconstruction routines.The construction of solid geometry, proposed by Oh et al. [47], and Yu and Zhang [18], involve the extrusion of recognized polygons by interacting with WIMP interface resources (mainly buttons). Naya et al. [49], Lopes et al. [50] and Kang et al. [43] propose supporting gestures for the sketch reconstruction control. Except for Masry and Lipson [19], preliminary performance simulations of building design are not explored.
Company et al. [45] and Naya et al. [49] offer high interoperability by using the DXF format, whereas Lopes et al. [50] offer high interoperability with other 3D formats. Kallio [29] provides interoperability through VRML files with generated data limited to 3D lines. Oh et al. [47] permit external performance simulation through exported models, while Dorsey et al. [44] allow the generation of perspective images, which cannot be used in subsequent design stages.
All the functionality principles mentioned should be integrated in one single software for an appropriated support to early stages of architectural design. Each of these functionality principles have a specific role or attribute needed in software for the early stages of architectural design and are synthetically presented in Figure 4 and described as follows.
Figure 4: Principles of functionality
needed in software for early stages of
architectural design.
4.1. Functionality principles
Multimodal, bimanual interaction: Interaction modes, inspired in traditional bimanual work—with stylus pen and multi-touch screens—define three interaction modes: (i) Creation: drawing and annotating with a stylus [49]; (ii)…
Related Documents
