Transformable structure in a responsive performing machine
The design project initiated with an interest in finding how the built environment affects human’s emotion and an idea of opening a conversation between the built surrounding and the state of mind of the person who inhabits in it.
In our daily life, people are constantly influenced by their surrounding even without knowing it and react in a subconscious or conscious way toward the environment. Everything continuously interacts with each other and there is some logic in the way they interact. It’s like working with a giant and complex algorithm, constantly calculating between input and output and present it with the combination of space and time.
However, generally speaking, the built surrounding would usually just passively being changed by the man’s doing acting upon it and present the result which serves as an input influencing the person’s feeling after receiving sense impressions.
We wonder, what if the stiff, lifeless and impersonal surrounding can actually react and interact with people by actively detecting a person’s emotion and responding to them instead of passively complying to people’s command. It would not be a building in any conventional sense but would be instead a socially interactive machine constantly translating and performing one’s state of mind in a creative way. (Mathews 2005)
Fig. 1, Design concept diagram
- Project content
The system of this design project can be mainly divided into two parts. One is measurement – the input, and another is performance – the output. The input part, the sensors, detect the player’s emotional reaction. After combine and analyze the data from different sensors. We perform it with spatial transformation via our deployable structure which is the output and the space the player situated in. At the same time, the shape-shifting structure will be presented in front of the player and consequently affect the person’s emotion. This will form an intriguing interaction loop between the person and the performing structure
- Measurement (Input)
Measurement is the part that detects and decrypt the player’s physiological changes which associated with emotion. Here we call it emotion-related biometric data. The biometric data include external and internal physiological reaction.
The external reaction we examine here is the facial expression, being one of the most informative and immediate ways for humans to communicate and express their feeling intentionally or unintentionally, will be detected by webcam and analyzed by Face OSC program. As to the internal reaction, we use galvanic skin response (GSR) sensor to measure human’s electrodermal activity (EDA) which is the property of the human body that causes continuous variation in the electrical characteristics of the skin. This biometric data can serve as an indication of psychological or physiological arousal.
After analyzing the meaning of the emotion-related biometric data received from the subject. We can generally distinguish whether the person was happy or sad, active or sleepy. As the fig. 2 show, basic emotion can be generally mapped by the horizontal axis which represents the level of positive emotion and the perpendicular axis which indicates the degree of activeness. (Russell 1980) The level of positive/ negative emotion can be inferred by the information we got from facial expression sensing system. The degree of activeness, on the other hand, can be detected by the galvanic skin response sensor. Therefore, we can now infer the subject’s basic emotion in principal.
Fig. 2. Eight affect concepts in a circular order by Russell, J. 1980.
How do we translate people’s emotion and to visualize such an abstract ideal by our installation? Firstly, we have to delve into the study of knowing how people perceive the space embedded in the surrounding. In The Journal of Transpersonal Psychology, written by John Welwood in 1977, he stated “A major dimension of feeling space is that of expansion and contraction. When feeling good, we are generally more expensive, moving outward, and able to take on more of the world without hesitation. In such an expanded space, we have a sense of large zones of uncharted territory surrounding us, as well as a sense of excitement about potential explorations to be made into these unfamiliar areas. At the other end of this dimension, when space contracts, it can feel as though we are being squeezed and crushed against our boundaries, and there is nowhere to expand. Our whole world feels compressed: the past is breathing down our neck, and the future is without promise, going nowhere. We may feel claustrophobia and panic, and want to burst through the walls that confine us, the barriers we are up against.” (Welwood 1977) As what he described, people usually have a certain feeling toward a certain type of volume transformation of the space they inhabit. Also, instead of a static and motionless object, a moving thing can draw more attention from the spectators and therefore have a larger effect on human’s emotion. Based on these researchers, we decided to try to capture and materialize the elusive concept of moods by building a structure which can perform expansion and contraction according to the player’s emotion-related biometric data.
- Performance (output)
In this design project, the deployable structure, the output of the whole system, which can expand and contract is the result we are going to present to the audiences. It is also the main focus of the thesis. Among a wide range of types of foldable structures. We found the suitable transformable structure through tried and tested methods which are the origami-inspired reconfigurable structure proposed by researchers from the Harvard University. It is capable of morphing its shape in three dimensions and creating an interesting visual effect by systematically alter the connected geometries between cube and polyhedron.
The term transformable structure, also known as deployable or kinetic structure, is generally used for a broad category of structures that are “capable of executing large configuration changes in an autonomous manner. (Liapi 2002) The configuration of such structures can be transformed from a closed compact configuration to a predetermined expanded form. (Gantes & Connor & Logcher 1994) It is widely accepted that the primary goal of transformable structures is to provide flexible adaptation to constantly changing needs, desires, and environmental conditions (Yiannoudes 2010)
The deployable structure can be perceived as an integration of two main components, i.e. the structural system and the actuation system since the deployment is particularly related to a morphological variation of the structure. (Grosso & Enrico & Basso 2012) Both components are essential but, while there are a handful of technical solutions can be used for actuation system, it is the former where the major conceptual design issues are to be confronted which is also the main focus in this paper.
- Historical background
Although deployable architectural elements and structures have existed since the ancient times and in different cultures, they were more widely recognized and developed after the Second World War due to the rapid changes in the western way of life. (Oungrinis 2009). Deployable lightweight structures and transformable, mobile or portable environments, built by architects and firms such as Buckminster Fuller, Hoberman associates and FTL Happold have sought to resolve economical, practical or ecological problems of the construction industry, and respond to issues of survival or nomadic dwelling. (Oungrinis 2006) On the other hand, the development of computers and cybernetic control systems inspired the design of more experimental transformable environments – such as Price’s Fun Palace, Archigram’s Living 1990 installation and Constant’s New Babylon – able to respond to change and individuality. Such visionary projects were pioneers of the so-called ‘intelligent environments’. (Yiannoudes 2010) More recently, the merging of kinetic architectural systems and digital technologies has produced digitally-driven kinetic architecture, structures or building components able to modify the shape, size or position of their physical form using embedded computational technology. This is a vision for technologically-enhanced architecture with ‘naturalised’ capacities – that is, sensing and actuation abilities, intelligence, and pro-active behavior. (Yiannoudes 2010) Although such applications are rather limited and exist mainly in experimental and academic contexts, there is indeed a growing interest in the potential development of digitally-driven kinetic architecture. As Michael Fox (2001, quoted in Yiannoudes 2010) of the Kinetic Design Group argues: “Architects need to design with an understanding of the current capabilities of embedded computation that have attained sufficient maturity to act as independent subsystems that can be beneficially incorporated into a kinetic design.”
A fundamental requirement to be defined as a transformable structure is to be able to alter the shape of its geometry. This requirement leads to the field of mechanism-like structure possessing kinematically indeterminate states which can also be called “Variable Geometry Structures (VGSs)” (Grosso & Enrico & Basso 2012) VGSs can be classified into four main groups according to their morphological and kinematic characteristics – spatial bar structures consisting of hinged bars, foldable plate structures containing hinged plates, tensegrity (strut-cable) structures and membrane structures. (Hanaor & Levy 2001) As the Fig. 3 show, the structures can be distinguished into two major categories. The first category – deformable – includes those that change their configuration based on the intrinsic property of their material such as compliant mechanisms, tensegrity and tensioned membrane structures, while the second category – rigid links – which usually alters their shape relying on the geometric inter-linking of their elements and contains amounts of essentially resistant bodies, which are connected by hinges employed to enable movement along on or more degree of freedom. Mutually supported elements, scissor-like, folded plates and morphing truss structures are among this category. (Hanaor & Levy 2001)
Fig. 3. Classification of VGSs on the basis on their morphological and kinematic characteristics by Hanaor A. and Levy R., 2001.
- Finding the suitable structure
According to what transformative effect we intend to achieve in our design project, which is spatial expansion and contraction, there are two kinds of preferable structures we have selected for project development: tensioned membrane structures and rigid foldable origami structures
- Tensioned Membrane Structures
Tensile membrane structures are spatial structures made out of tensioned membranes which carry only tension and no compression or bending. Tensile structures are the most common type of thin-shell structures. In this kind of structure, membranes work together with cables, columns and other construction members to construct a form. A tensile membrane structure is most often used as a roof, as they can span large distances economically.
- The primary reasons we choose this form of structure:
- Their soft and streamline appearance seems more approachable and natural.
- Can economically create a more dramatic visual effect with the ability to span large distance, especially when vast open spaces have to be covered.
- The low weight of the materials also makes construction easier and cheaper than standard designs.
- Prototype of Tensioned Membrane Structure
The concept of this prototype is to perform human’s facial expression and heart rate with the tensile membrane structure by the deformation of the elastic fabric fixed on the top of the device. We anticipated this structure to scale up to enable people to walk in and interact with it. With the visualization of the player’s facial expression and heart rate, we intended to create an empathetic experience among the audience.
Fig. 4. Sketch of design concept-1 Fig. 5. Sketch of design concept-2
The prototype was made with clear acrylic, white elastic fabric, white threads, servos, and Arduino. There are X major part of the device: 6 servos, Arduino, heart rate sensor, acrylic frame which hold the fabric, filters and servos, two layers of filters with the upper layer keeping the thread straight and the lower layer arranging which threads should be controlled by the certain servo and the membrane made with elastic fabric which is fixed on the frame with its four sides and tied up with 28 threads.
Fig. 6. Membrane structure design prototype – 1 Fig. 7. Membrane structure design prototype – 2 (heart rate sensor)
- Folded plates structures
Folded plates structures are inspired by Origami, the Japanese ancient art of paper folding, providing an ideal platform for the design of transformable systems with its astonishing formal richness and variability. Based on a simple technique, a myriad of shapes can be achieved by folding paper along pre-defined creases. By substituting the paper and creases with rigid panels and hinges respectively, this technique can be used to develop some special mechanisms of deployable structure in the applications of transformable architecture. (Hanaor & Levy 2001)
- Origami moving cubes
At the early stage of this research, inspired by origami moving cubes (Fig. 8), we’ve designed a type of structure comprising 24 cubes and 8 tubes which can be seen in Fig. 9. This configuration contains three layers of cubes and two layers of tubes with the same length of each side. AS shown in Fig. 9, the layers of the cube can transform between two states by varying the angleα1 from 0 to π and trigger the deployment of the whole system while each layer of the tube can move separately between three states defined by the angleα2 and α2 altering between 0 to π. This structure can alter its shape partially and have rich mobility. However, the transformation is not the deployment form we attempt to achieve which is contraction and expansion.
Fig. 8. Origami moving cubes Fig. 9. Analysis of the transformation of origami moving cubes
Fig. 10. Analysis of the transformation of origami moving cubes – prototype
- The origami-inspired reconfigurable structure
While most of the origami-inspired design projects rely on two-dimensional folding patterns, such as the sound-shaping mechanical ceiling project (showed as fig. 11) proposed by the design firm RVTR, the research group led by Overvelde, J.T. et al. (2016a) from Harvard University (Soft Robotic Matter) introduced an innovative method to design three-dimensional reconfigurable structures consisting of a periodic assembly of rigid plates and elastic hinges. This robust design strategy based on space-filling tessellations of polyhedral is inspired by the structural diversity and foldability of the prismatic geometries constructed using snapology origami and modular origami technique. This approach can be widely applied to everything from meter-scale architectures to nanometer-scale photonic systems. In order to find the suitable transformable structure for our design project, we built a series of physical folded paper models and computational virtual models to explore the formal and spatial potential of this technique.
Fig. 11, Sound-shaping mechanical ceiling project design by IAAC students Shambayati, R., Tankal, E., and Baseta, E. (2016)
- Study background
Snapology, a type of modular united-based origami, is a paper-folding technique that was discovered by Heinz Strobl. It involves using long strips of paper to create complex geometric extruded polyhedral. (Fig 12) However, some of the extruded polyhedral geometries are stiff and almost rigid such as the extruded icosahedron shown in Fig. 12 while others have multiple degrees of freedom and can be easily deformed as the Fig. 13 shows. With an aim to develop transformable structures, we have chosen a selection of foldable truncated polyhedra made by snapoloy technique based on the research of Soft Robotic Matter group from Harvard as templates to analytically and experimentally study their mobility and how do the shape and volume of the structures’ inner space change after assembling the unit cells into a modular structure.
Fig. 12. Example of snapology by Overvelde, J.T. et al. (2016a)
Fig.13. Extruded cube by Overvelde, J.T. et al. (2016a)
The idea of constructing the larger structure by assembling several arrays of identical geometric unit cells is inspired by another origami technique – modular origami also called unit origami. These three-dimensional paper art are created from a number of small pieces of paper that are easily folded and then cleverly fit together to form a spectacular structure. The shapes range from polyhedra to bristling buckyballs. The number of units used, and the way the units are assembled can change the appearance of the final model dramatically. (Hart 2017)
Fig. 14, Mind-Blowing Modular Origami by Hart, G. (2017)
- Study approaches
- Physical prototype building experiment
The fabrication approach we use to build the prototypes at the stage of the experiment is developed by Overvelde, J.T. et al. (2016a). It is fabricated with two layers of cardboard with a thickness of 0.5 mm and one layer of double-sided tape with a thickness of 0.1 mm in the middle, using a stepwise layering and laser cutting technique on a CO2 laser cutting machine. The layers were cut in three steps to form flat building blocks with both flexible and rigid regions. (Overvelde, J.T. et al. 2017) (Fig. 15)
Fig. 15, Fabrication approach of extruded cube by Overvelde, J.T. et al. 2017
- Computational assistance and simulation
In order to build the modular assemblies and study the transformation of the shape and inner-volume of the complex geometric configuration structures more effectively and accurately, we used Rhinoceros and Grasshopper software in our research process. Grasshopper is a visual programming language that runs as a plugin within the Rhinoceros 3D computer-aided design software. It is a graphical algorithm editor which takes advantage of Rhinoceros’s existing tools and provides new ways to expand the 3D modeling processes such as creating 3D models through mathematical functions, automating repetitive processes, effortlessly constructing and making changes to complex forms through repetitions of geometry, allowing designers a high degree of flexibility in creating virtual three-dimensional models. In this paper, we use this technology to construct complex configuration, simulate its potential of transformation, help find the suitable arrangement according to different unit cells and analyze the change of the inner space’s volume and coverage.
Fig. 16, Computational assistant – Grasshopper in Rhinoceros
- Extruded cube
- Unit cell study
Starting from the extruded cube as the fundamental unit cell shown in Figure 17, they construct the unit cell by removing all the faces and extruding the edges of a cube in the direction normal to each face (see Fig. 17). The result is a 3D structure with 24 faces connected by 36 edges of length L. If we assume that all the faces are rigid and the plates can only fold the edges, such a periodic structure will have three degrees of freedom identified by the angle α1, α2, and α3. (Overvelde, J.T. et al. 2016a) Changing these three angles not only deforms the assembly of plates into numerous particular states of shape but also significantly alters the volume of the inner space defined by them. To find an ideal structure which can better perform contraction and expansion, it is essential to analyze how the shape and volume of the space change inside the geometry and the structure built with several arrays of this geometric unit cells. To describe them, we use V i to stand for the volume and A, B, C, D, E, F, G, H as the eight vertices of the fundamental rhombohedron inside the extruded geometry and introduce the vectors P1, P2, P3 spanning the internal rhombohedron. (Fig. 18) When (α1, α2, α3) = (π/2, π/2, π/2) (in Fig. 18, it is shown as state #1), the structure is fully expanded with the maximum Volume: V i = L3.
Fig. 17. Extruded cube unit cell by Johannes T. B. and his research group, 2016
Fig. 18. Analysis of the transformation of extruded cube
To transform from state #1 to state #2, it can be achieved by three different directions of movement of rhombi ABCD (here we call it RABCD) and rhombi DCGH (RDCGH) which are -P3+ P1, -P3+ P2, for QABCD and P1+ P2 for QDCGH changing the shape into state #2-1, #2-2 and #2-3 as a cuboid with two of the six extruded rhombi flatten while (α1, α2, α3) = (0, π/2, π/2), (α1, α2, α3) = (π/2, 0, π/2), (α1, α2, α3) = (π/2, π/2, 0) respectively. As to state #3, the movement is 1/2 P1 for RABFE + -1/2 P1 + P2 for RABCD to form a hexagonal prism when the angles are (α1, α2, α3) = (π/3, 2π/3, π/3). The last one is fully folded into a plate by the movement of -P3+ P1 for RABCD + P1+ P2 for RDCGH when (α1, α2, α3) = (0, π, π). (Fig. 18)
The transformation of this geometry was simulated by grasshopper as shown in Fig. 19. The motion of this extruded cube is defined by three angles – α1, α2, α3, which can be altered manually by dragging the three sliders namedα1, α2, and α3. This function can be extremely useful when it comes to building complex modular configuration in the later experiment.
Fig. 19. The transformation simulated by Grasshopper
- Assemble the modular structure
After the study of the single extruded cube units, the next step is to build up the structure by assembling the basic unit cells systematically. Before building the assembly, we need to understand how it can be connected and how each connected building block move or rotate. In Fig. 20, it shows that the extruded cube unit cell has six tubes open to six direction – front, back, left, right, up, and down. We can connect six identical unit geometries on its outer edges and study its transformation from state #1 to state #4. It is clear that unlike the geometry we will talk about later, there is no limitation in terms of the direction of each module’s arrangement here. This is because the extruded cube unit has a highly symmetric structure which increases its flexibility but can only perform a rather predictable transformation.
Fig. 20. The connection of the extruded cubes at different states
As can be seen in Fig. 21, we attached the outer edges of 64 extruded cubic unit cells to construct a 4 x 4 x 4 cubic framework and took out 8 of them located in the center. The purple area represents the inner space of the built shell-like structure. The picture indicates that the assembly still has three degrees of freedom and deforms in exactly the same manner as each constituent module. By altering the microstructure (that is, the unit cells) into the various possible configurations described in Fig. 18, the macrostructure of the assembly can be significantly changed with its initial cubic shape deforming either into an extruded rhombohedron, a hexagon, or even a completely flat 2D state.
Fig. 21. Analysis of the cubic modular structure
Generally speaking, the inner space, (as the purple area shown in Fig. 21) transforms from a cube to a diamond column with the height remains the same from state #1 to state #2, which means it is squeezed in horizontal dimension with a more contracting percentage on one direction. As from state #2 to state #3, space morphs into a twisted rhombohedron and becomes much shorter and can even completely be folded into layers of flat plates at state #4. The volumes of the inner spaces are 512 L3, 120L3, 24 √3 L3, and 0 at state #1, state #2, state #3, and state #4 respectively. Additionally, it also indicated in Fig. 21 that the areas covered by the outer unit cell are 144 L2, 72 L2, and 56 L2 from state #1 to state #3 with the opened areas being 240 L2, 120 L2, and 12 √3 L2 respectively. Therefore, the shrinkage ratio from state #1 to state #2 and state #3 will be 23.5 % and 8.12 % respectively, while the coverage of the shell structure will surge from 37.5 % at state #1 and #2, and rise to 73 % at state #3.
- Extruded hexagonal prism
Next, we tried to extrude the faces of a hexagonal prism as shown in Fig. 22. Compare to previous structure, this model has more degree of freedom, meaning it can deploy into more states of shape. However, the shape transformation in the internal space is quite similar to the extruded cube so that we decided to develop other geometries instead.
Fig. 22. The transformation of extruded hexagonal prism
- Extruded truncated octahedron
- Unit cell study – Initial prototype
The third geometry we examined was a truncated octahedron with eight hexagonal faces extruded (shown in Fig. 14), four of the square faces being removed and two of them being remained since the structure will be rigid if all the faces are extruded. This resulting structure can be transformed into three distinct configurations by varying the angle θ. (Fig. 14) By altering θ between 0 andπ/2, the configuration of the system can be altered into state #1.2 and state #1.2 when the inner space being laterally squeezed or stretched in the vertical direction and contracted in the horizontal direction as shown in Fig. 14. The volume alteration of the internal truncated octahedron between each state can also be seen in Fig. 14 as 4L3, 8 √2 L3, and 0 for state #1 to state #2.
Fig. 23. Analysis of the transformation of extruded truncated octahedron with 8 faces extruded
- Unit cell study – Enhance the transformability
Based on the truncated octahedron, we tried to construct another similar model but extruded only six of the hexagonal faces and make the rest of the faces rigid. The purpose of this modification is that it can reduce the connectivity and so as the stiffness when we connect each unit on the extruded edges to assemble a periodic tessellation of convex polyhedra. This approach is based on Johannes’s research where he discovered that higher values of average connectivity of the unit cell which defined by the number of extruded faces of the polyhedral lead to a more rigid configuration. The resulting two different states of this transformation can be seen in Fig. 24. As shown in this figure, the geometry alters its shape according to the changes of angleθ. At state #1, θ=π and the shape of the space inside is a hexagonal prism with its volume equals to (9 √3/2)L3, while the volume reduces to 0 whenθ=π/3 after expands to its maximum, 8 √2 L3, at state #1.2 with the angleθ= 2π/3. Thus, the volume increases to its 144.87 % from state #1 to state #1.2 and declines to 0 at state #2. In terms of the shape transformation, this unit model can perform a much preferable motion of contraction and expansion as the movement’s direction of each plate is either toward or outward its center point.
Fig. 24. Analysis of the transformation of extruded truncated octahedron with 6 faces extruded
Fig. 25. The deployment of octahedral module defined by the angleθ
- Assemble the modular structure
To assemble a multi-unit configuration of an extruded truncated octahedron, there is a noteworthy limitation of its direction of each connected feet. This feature is formed by two different layers of six extruded feet face outward from the center with three of them higher than the rest which is highlighted in red color in Fig. 26. As the shape transforms from state #1 to state #2, these three upper legs turn up from three horizontal directions while the lower feet are turning down. It is clearer to demonstrate it in Fig. 27, which shows two different ways of connection: mode A and mode B and illustrates the unit cell’s deployment form two states. In mode A, the two modules are attached by feet of a different layer, i.e., the upper leg and the lower leg. When it alters to state #2, the two unit geometries stagger from up-left to down-right which can avoid motion conflict after attaching more unit cells. On the other hand, the configuration built by two units with feet of the same layer connected will inevitably face some limitation in terms of extending its structure as shown in mode B in Fig. 27.
Fig. 26. The directions of the extruded tubes
Fig. 27. The limitation of connection
Therefore, we attached each module with identical direction and found that its extensive structure will grow outboard linearly from the center with three of them slightly downward and three of them marginally upward. (Fig. 28) This angular differentiation will become even larger after the assembly morphs into state #2 as shown in Fig. 28.
Fig. 28. The positions of each unit from state #1 to state #2
We fabricated a prototype containing 8 units of extruded truncated octahedron to test its deployment and discover that in terms of the stability of the configuration, which is also crucial for building strong structure capable of performing smooth transformation and being more compatible with the pneumatic actuator, the test results suggest a network of structure with each unit joined together to form several rings will be far more strong and easy to transform than each linear branch developing separately.
Fig. 29. The physical prototype of extruded truncated octahedron
Thus, we constructed another shell structure with a network of repetitive geometries and explored its shape-shifting ability as shown in Fig. 30. In order to compare this extruded truncated octahedral structure and the extruded cubic assembly, we created each unit with the same length of edges, L, to form a 4 x 4 x 4 structure just like how we built previously with the extruded cubes. (Fig. 30) The resulting structure was similar to two hexagonal pyramids being combined together upside down. Unlike what we have discovered with the single octahedral unit cell, the lowest volume of the space inside the assembly appeared at state #1 and almost enlarged to its twofold at state #2 with the figure being 453 L3, and 771 L3 respectively. However, the highest volume, 1123 L3, can still be found at state #1.2. As to the area of the inner space, the figure was quite similar as being 873.5 L2 and 743.5 L2 at state #1.2 and state #2, while the area reduced to only half at state #1. Lastly, the coverage rate of each state was rather the same, being around 54% to 65%.
Fig. 30. Analysis of the assemble with 64 extruded truncated octahedral unit cells
Fig. 31. Modular structure comparison
As the figures are shown in Fig. 31, different geometric modular structures can have drastically diverse characteristics even built with the same length of edges. The greatest volume of extruded cubic assembly’s inner space is even less than extruded truncate octahedral modular structure’s half, while it can significantly expand or shrink between 0 to 512L2. It is also noteworthy that the coverage rate stays the same from state #1 to state #2 and rise to more than its two times percentage for the cubic assembly. Considering the octahedral structure’s limitation and a rather stiff deployment. It seems that extruded cubic modular configuration will be the better choice to form a more complex structure and perform a more dramatic transformation.
Although it was only in the early 20th century that architects began to widely discuss the possibility for movement to be enabled for a significant portion of a buildings’ structure, a various form of kinetic architectural systems have been developed and established. However, there is still tremendous potential for the new evolutional structural technique to chart the unmapped territory. A new class of three-dimensional reconfigurable structure inspired by snapology origami technique has just been proposed recently, yet remained undeveloped in the application of deployable architecture.
In this paper, during the process of searching for an ideal transformable structure to build an interactive performative pavilion, we discovered this innovative structural technique. We found the suitable module through analyzing the form of the movement of the structural transformation and calculated the volumes and coverage rates of the space inside the assembled configuration. Ultimately, we integrate it with pneumatic actuation system to breathe the prismatic creature into life. The result is satisfying yet there are still some obstacles have to overcome. In order to fight against the gravity and lift the weight of the structure by providing more power and acting more effectively, the improvement should be made on the actuation system. In the future, it can be helpful to combine air muscle system with our current actuation system and rearrange the position of the air pockets. Furthermore, we will also refine the design of air pocket and show the actuation system by transforming the plates into a frame and combining them as bones and flesh. The configuration of the templates can be an integration of modules in various scales and in the end, create an interesting performative shape-shifting installation.