Framing responsive architecture with soft robots—the exploratory practice of soft pneumatic robotic architectural system

Inspired by intelligent skin, this research constructs a soft pneumatic robotic architectural system based on the interaction between inflatable material agents and the human body. It aims to provide flexibly changeable scene modes for dynamic inhabiting through responsive spatial performance. From the idea of the Fun Palace and the structure and performance spatial methodology to the proposal of responsive/soft architecture and relevant examples such as Furl and Robot Soft, the literature review provides the foundation and direction for the establishment of research experiments. The research methodology is based on structure and performance, oriented to explore soft material agents, link input information and output modules to realize the manifestation of human-space interaction and spatial performance and assemble the basic configuration of the system. As a result, a synergistic architectural interaction system like a media center that can provide a variety of adaptive space scenes is built, which can realize real-time perception and responsive control of the interactive process, and at the same time provide richness of scene performance beyond established architectural configurations, confirming the possibility of creating a behaviorally responsive architecture by integrating modularly morphing spaces, AI-sensing systems, and interaction technologies. This study explored responsive human settlements by expanding soft pneumatic robots in architecture. It can expand to fields such as machine cognition at the input level, further realize the real-time feedback system at the output level, and can evolve with the information interconnection of urban landscapes and residential communities in terms of spatial interaction. The work gives a convincing vision of architectural intelligence through soft robotics and is expected to continue to develop and bring more inspiring innovations with its integrated methods and connotations.

place, and express its function at a particular occasion.Besides, with the accelerated pace of modern life and the rapid development of new technologies, architecture faces the challenge of rapidly changing scenarios over time.In the new era, architecture should be created as a responsive and performative structure enriched with complete resources and more possibilities for civic lives.Meanwhile, as a coworking lifestyle emerged with working and socializing being more flexible in place and manner, people especially the young, preferred to live in open shared places.The SPRS System is built to search for the solution to this situation.
To unleash the possibility of future architecture but keep the research on an architectural scale, the system is set as a home and workplace, and it is constructed to create a hybrid shared space and life-working dynamics by space scene changing in the living and working place of freelancers and other life-work balancers.
To manifest the target of the research problem, Some survey and user research about the coworking living space is conducted.Usual working environments, dissatisfaction and expectations for the workplace, and expectations in the life of the users are focused on inspection.From the outcomes, it is found that the users prefer more adaptable functions and some personal areas, they need office space to be more organized, and shared space should encourage communication.It guides design goals to space-saving, utilization improvement, order rebuilding of work, life, and social activities, and creating new possibilities for working and career (Fig. 1).
In the 2018 China House Vision exhibition hosted by Kenya Hara, the Cross-boundaries design team Fig. 1 User persona and survey outcomes based on the user research about the co-working living space combined with smart home design to design a personal space called Infinite Living, which inspired SPRS research inception.It introduced an intelligent skin implanted into spatial structures, which can actively adjust the spatial form and scene according to the daily living conditions, behavioral patterns, and emotional changes of residents, as well as seasonal needs in nature and the atmosphere.It is meant to re-enable all restricted spaces, which can be galleries, clinics, restaurants, banks, etc.It is implanted into the space floor, wall, and other structures to cover almost all space surfaces, the space partition is removed, and the original windows are replaced by completely intelligent glass walls.The intelligent skin can freely switch between working mode, sports mode, entertainment mode, and communication mode, creating suitable space scenes and providing necessary space elements.In the case of usage scenarios and space perception, the spatial skin will make active adjustments, optimize the experiences, allocate space resources rationally, and realize interactive spaces (Fig. 2).
The intelligent skin gives a vision of deformed architecture that can be refined through a technological blueprint.Simultaneously, the emergence of interactive technology and new material prompted traditional architecture via intelligent transformation in terms of morphological property.Of note, the invention of robotic soft architecture has made the morphological shift and its cybernetic formula possible.Numerous research cases proved that the outstanding paradigm as a soft responsive architecture via technical, functional, and actuated directions had been endowed.This paper focuses the main research problem on the intelligent dynamic living pattern, which has become increasingly prominent because the impact of digitalization and mobility development on space use requires us to explore the transformation of architectural patterns.In response to the problem, a set of criteria for developing a scheme of the soft robotic architectural system focusing on responsiveness and performance, based on which, relevant components are created to utilize the space interaction with people through morphology-variating modules and body tracking signals; with flexibility increased significantly by the incorporation of deformed components, the combination of two mechanistically synergistic circuits formed an interactive architecture system.The cybernetic architectural system introduces the intelligent traction methods that buildings can link, in which the intelligent systems and interfaces designed will serve as examples and clues to provide means and perspectives for the application of artificial intelligence to affect the built environment.

Literature review: implementation of responsive architecture in soft robotics
The previous architectural research, technological practices, and social proposals mentioned the initial architectural transformation by shifting shapes only.Essentially, we focus on the structures that live according to the changes in their environment.They breathe differently in different conditions and make changes in accordance with the environment (Lundén & Kauste, 2018).Thus, they may have a little bit of their own life.The original idea considers something close to an animal, namely about re-establishing a relationship with architecture.Among the explorations of the forerunners of this kind of soft architecture, such as Archigram, Buckminster Fuller, and Yona Friedman (Bhatia, 2013), the Fun Palace conceived by Cedric Price in 1964 was almost taken into practice.While developing his design for the Fun Palace, he described his visions for such a place (Mathews, 2006): "Old systems of learning are now decayed.The new universities will be of the world and in each man… The variety of activities cannot be completely forecasted; as new techniques and ideas arise they will be tried.The structures themselves will be capable of changes, renewal, and destruction.If any activity defeats its purpose, it will be changed (Fig. 3)." Cedric Price pointed out that in allowing for change and flexibility, it is essential that the variation provided does not impose a discipline that may only be valid at the time of design; It is easier to allow for individual flexibility than organizational change -the expandable house; the multi-use of fixed volumes; the transportable controlled environment (Jencks & Karl, 2006).In this way, the concept of improvisational architecture was introduced through an entity, whose essence was in a continuous process of construction, dismantling, and reassembling by permitting multiple and indeterminate uses.Gorden Pask, the technical consultant of the Fun Palace committee, emphasized the functions performed by human beings or human societies and claimed that a building cannot be viewed simply in isolation but as a human environment.It perpetually interacts with its inhabitants by serving them and controlling their behaviors (Pask, 1969).New correlations in terms of the crucial relation of structure with its target were developed with another theory by Herman Hertzberger.He introduced the linguistic concepts of 'competence' and 'performance' addressed by Chomsky to illustrate the essence of architecture (Hertzberger, 2005): competence is the knowledge that a person has of his or her language, while performance refers to the use he or she makes of that knowledge in concrete situations; It can indeed be established with architecture, In architectural terms, competence defines a form's capacity to be interpreted, and performance, stated by the form, interprets a specific situation (Fig. 4).
All previous relevant research advocates a new kind of active and dynamic architecture to permit multiple uses and constantly adapt to change.It can be a network of diverse events and a space of oscillation between incongruous activities simultaneously.To redefine the users, the architectural space, and scenes adapting to different functions and interactions of humans with space, a creative and flexible control system is needed by the users in a dynamic space pattern for various daily activities.New technology such as artificial intelligence and cybernetics is expected to be followed.In addition, the research on material systems and soft robotics in search of an alternative to rigidity can be accepted.As Nicholas Negroponte proposed that responsive architecture or soft architecture machines during spatial design problems can be explored by applying cybernetics to architecture (Negroponte, 1975).Meanwhile, by forming a new frontier of kinetic design, the domain of soft robotics has created an exciting and highly interdisciplinary paradigm in engineering, which provides a method for revolutionizing the role of robotics in architecture.
To date, many studies have explored the application of soft robotics to architecture from various dimensions.For example, researchers of the Interactive Architecture Lab designed a soft pneumatic pavilion called "Furl" with cybernetic silicone-made components (Mangion & Zhang, 2014)."Furl combines Electroencephalography (EEG) with advances in soft silicone casting of "air muscles".The introduction of soft robotics replaces the mechanical principles in interactive architecture through a biological paradigm.EEG allows sensing of human brain functioning so that the environments begin moving and responding to users' very thoughts; The designed components have a wide palette of deformation patterns of inflation.Through the combination of soft and hard architectural elements, "Furl" creates a new platform for a kinetically responsive architecture that can let space interact with users' needs and adapt itself to environmental conditions (Fig. 5).
Compared with "Furl" which focuses more on its adaption of architectural structure, Robot No. 6 "Soft: Balancing and Wiggling" by researchers from Princeton University mainly worked on robotic responsiveness and the controlling mechanism (Shi, 2018).It is a pneumatic architectural robot with polyurethane airbags and 3D-printed joints controlled by homemade inflate-andbalance algorithms.The core function is to use pressurized airbags to lift the weight.The sensing-feedback algorithm prevents the whole thing from falling by separately controlling the pressure in different airbags.This prototype can be applied to several real-life situations.It can work as a critical component for an emergency shelter for its lightweight and fast deployment or as an additional interface between humans and buildings that is safe and responsive (Fig. 6).
Challenging conventional design thinking about adaptive architecture, these experiments outlined suggest approaches to building a soft responsive architecture.Furl focuses on the response mechanism, while Robot Soft goes deep into the pneumatic system, both of which provide the basis for the interactive input and output vectors of experimental design.Soft responsive architecture acts as a proxy for the improvisational performance of architecture, enabling the theoretical visions of deformed architecture through technology.Through it, users and designers are in complete control of the continuous communication with the machine in the entire process.In this human-computer symbiosis, architecture became a mechanism of information exchange, moving beyond computer pre-configuration and pre-formulation.Considering the initial experimental goal of releasing the possibility of building, the background setting is the working and living space, and the needs of multi-activity space and scene conversion should be met according to the user portrait.Besides, the referenced projects have their own limitations while providing the basis, thus providing clues to the focus of the experiment: The Fun Palace project inspiringly advocates possibility and feedback mechanisms, despite the unknown may lead to the excessive physical and  computational capacity required for the construction; Herman's spatial methodology provides an overwhelming theoretical basis for the experiment, while its specific realization of the artificial carrier requires more technical and research support; The Furl gives a good focus on the sensory interaction level and realizes space interaction, but it still needs to be more connected with the scene and placed in the specific and long-term use of the space; Robot Soft creates a specific and complete mechanism of action and feedback firmly in technology and structure, however, it needs to consider more elements based on the essence of architecture than that.
The above-related literature review provides the theoretical origin and technical basis of soft architectural robots and also guides the direction of future research we need to do from various aspects.It is not new for robots to intervene in the field of buildings, but we may need to pay more attention to how and what kind of results it leads to in construction.Meanwhile, although the soft robotic architecture is technologically viable, the inconvenience in manufacturing and excessive variates need further development to accommodate the popularization and diversification of space use; One way of empowering diversity is to standardize the intelligent components with varying soft elements in specific-sized modular and freely combine them for diverse spaces, and the generated results lead to various specific architectural use scenarios.

Methodology based on structure and performance
To provide a shared space of hybrid functions satisfying the compound responsive structure, this research is mainly focused on the adjustability, behavioral orientation, and performative aspect of the architectural system.Spaces have been considered introverted or extroverted atmospheres by categorizing them depending on the diversity of activities such as connected vs closed, private vs open, and stable vs dynamic.To determine one's expectations for improvisational space, some interviews were performed to summarize several types of scenes in different modes.However, the list was not comprehensive, as the variety of activities could never be precisely forecast (Fig. 7).

The exploration of material agent
Typically, soft robotics contains at least four parts: multimaterial fluidic actuators, modeling of soft actuators, sensing and control, and translational applications.Creating artificial mollusks not only develops a powerful, flexible material but figures out how to precisely control and cleverly manufacture it.One of the major challenges Fig. 7 Space scenarios adaptive to multiple activities in soft robotics is finding the right material for motion and sensing.When selecting the material as the system agency, hardness and deformability were the first two prime factors.Stiffness is related to spatial properties, while deformability affects the flexibility of spatial transformations.On the balance of the architectural hardness and soft deformability, in this study, inflatable silicone as the material agent is chosen, and it is combined with the hard plate as the architectural component (Fig. 8).These programmable materials provide a vision of a future where "programmable matter" and "computational materials" shed their technology-centric prefixes and become simply just "materials" (Tome, 2015).

Inputs, outputs, and interaction
The responsive approach to any architecture design affects certain environmental conditions or users' needs in terms of simulative responses to them.This kind of system works with two components, namely, the sensors, which are the input, and the actuators, which are the output signal.The sensors measure real-time data such as light, temperature, humidity, movement, position, and speed.These data are fed to the system to trigger the actuator, which performs in terms of changes in its shape, color, size, position, and geometry.The main medium of architecture in this cybernetic system was set as the basic elements such as the smart floor, ceiling, and walls.As the proxy of the morphological changes of the architecture, the soft modules generate outputs to render the performative outcomes of the different activities, a unique atmosphere, and separate privacy.In this experiment, Level 2, the human-system interaction, was further investigated.The IMU sensor was used to record the tilt angle of the user's body as input.The human body posture captured by the IMU sensor is defined through coding to the modular inflatable structure.The combinations of different modules and their changes proceeded to create a lively scene suitable for different functions and a flexible change of life mode.
The movements are represented on space curves coming from the phase space, which records the axes of joint angles and torso location and attitude.The system uses a multi-class statistical model of shape to obtain a 3D representation of the head and hands in a wide range of viewing conditions for tracking and interpretation of people.It makes the body gestures of the users easier to capture and recognize by sensing devices and AI computers (Fig. 9).The perceiving system is set to run on an intelligent computer standardized with a video camera and performs stably on people with different physical locations.It uses a multi-class statistical model of shape to segment users from a background scene and find and track people's heads and limbs in various situations.It can produce a real-time re-presentation of the users for the wireless interfaces, video databases, and coding of the system.Research by Lee Campbell and Aaron Bobick on the recognition of human body motion (Campbell & Bobick, 1995) and by Christopher R. Wren on real-time tracking of the human body provide inspiration and a foundation for the sensing system setting (Wren et al., 1996).
The current investigations involved the construction of an interactive system and its sampling by analyzing combined morphological models in diverse predesigned patterns.To realize the space change of a scene transformation, component modules were elaborately designed in specific patterns consisting of different calculated material attributes.The components inflate into conditional shapes to meet the requirements of various scenes.The tilt angles of the body along 2 axes of the IMU sensor using 3 left and right arm positions were used as the input data (Fig. 10), and the inflatable structure modulating in 3 m* 3 m as the material agent was measured for the space performance output (Fig. 11).This experiment's modules are conducted at a ratio of 1:30.The silicone material was selected for its elasticity to conform to the cambered surface, durability, and adaptability for flexible arrangement, and it can also be forged tightly with other materials.The assembled architectural system is displayed as a soft pneumatic robotic structure (SPRS) with silicone inflatables on 3D-printed acrylic plates controlled by mechanical-inflatable algorithms.In addition, a set of tracheas, valves, manifolds, adapters, connectors, circuits, Arduino boards, wires, etc., is formed in the system.

Fabrication and systemization
Thus far, the fabrication research of soft modules includes several aspects to be focused on.The manufacturing procedure faces the most duration for the casting complement of the rubber after blending two raw materials, since the inadequate time may not allow the downright weld of the weak joints, which leaks the gas and results in the inaccuracy of the experiments.The ratio of the raw materials also has a sequential effect on the toughness of the finished rubber, which directly influences the swelling capability of the soft samples.The silicone modules were connected to the inflator bump to generate the expected inflation, controlling the air pressure to the modules in graded amounts.This method obtains the diverse surface variations of the materials.Although two samples were noticed at risk of breaking the elastic extent, leading to abnormal morphology at the highest level of air compression, the variations with respect to others are negligible.In addition, the results confirm that the different morphologies are the consequence of the inconsistencies in the pattern, and their curvature rates and inflation sizes are obtained under elaborate manipulation.The air pressure in the module determines the volume of the morphology of a specific shape and the bending orientation depends on the main direction of the actuating material.Meanwhile, the combination of multiple samples created flexible abundant shapes of space, and it is also testimony to the necessity of the initial design of variable module patterns.As one part of the pneumatic interaction system, SPRS combines human-capturing censoring and is developed through space monitoring manifestation (Fig. 12).

Results as a synergistic system
All the material, interaction, and fabrication studies are integrated into an interactive synergistic control system.Importantly, SPRS is always dynamic, in construction, changing the morphology with time.The new relationship between SPRS, "the architecture", and its architecting gives us new thoughts on the future of architecture.From a broader perspective, it is more like a convergent system that incorporates mechanics, electronics, graphics, learning, gesture capturing, and vocal commands, a "media" center (Molly Wright Steenson, 2017).
The control system SPRS includes two parts: the electrical circuit composed of Arduino, IMU sensor, solenoid valve, and relay, and the air pipeline consisting of the solenoid valve, air compressor, inflation module, etc. Figure 8 shows the structural results obtained from it.It performs robust actuation in desired features by the digital control of human interaction and the SPRS (Fig. 13).
The current input setting is based on tangible interfaces and physical interaction, which is primarily to build up the fundamental logic of the interaction workflows; In regard to the upper layer of the sensing framework that introduces artificial intelligence such as machine learning, the scene-switching mechanism can match human activities in real-time by inserting the body tracking data captured by the AI sensing devices into the cybernetic interaction process (Fig. 14).
The samples are dimension-adaptive for the common functions and scenarios, providing a precise exploration of the pattern design of the modules (Fig. 15).
A quantitative analysis was applied to determine the swelling behavior based on the maximum curvature of the calculated levels within the considered position and scale (Table 1).

Fig. 11 Module configuration
The full volume of the "air compressor" is 500 ml, and at normal temperature and pressure, the molar volume of the gas is as follows: Thus, the number of moles is equal to the volume of the gas divided by the molar volume of the gas: Because T = 293 K and R = 8.3 J/K, according to the ideal gas law The pressures of module No.1 in the three modes of Pressure I, Pressure II, and Pressure III are calculated as: Similarly, all the internal pressure information of the inflatable modules can be obtained (Table 2).
Under the same amount of inflation, different patterns lead to different shapes, volumes, and pressures, thereby generating different spatial shapes, sizes, and hardness.Among them, Module 1 and Module 4 have the largest volume changes, which are suitable for responding to large space changes; Module 3 and Module 5 are the flattest, suitable for close-to-plane spaces such as tables, beds, raised floors, or raised walls.Module 2 has the highest height and is suitable for spatial changes that extend in one direction; Module 6 has undulating shapes and the highest pressure and is suitable for creating spaces with high hardness or height difference; Module 1 and Module 3 have the lowest pressure and are suitable for soft architectural scenes.Additionally, it should be noted that the pressure change trend of the variable module is not the same under different levels of inflation (within the appropriate range).The pressure of Module 2 and Module 3 decreases with the increase of the inflation level, and the pressure of Module 5 and Module 6 As the inflation level increases, the pressure of Module 1 and Module 4 first increases and then decreases with the inflation leveling.Therefore, the selection of a specific inflation volume should also be considered appropriate.
Based on this approach, the prediction for the average uncertainty of the model in this study slightly exceeds the acceptability limit defined by the previous research.Nevertheless, these results suggest that data obtained using SPRS to simulate material inflation and space construction can provide more information for assessing the impact of performative strategies than that of the traditional setting of collaging configuration.Of note, it can be observed that the intermediate zone created by the arrangement and combination of the diverse samples is partly unexpected.There is a striking richness noted when each of the modules forms a synergic entity of the morphological space and structural system in the adaptive intelligent SPRS for physical interaction (Fig. 16).

Discussion toward next work and architectural future
In terms of constant change, impermanence, process orientation, and interchangeability aspects in improvisational and responsive architecture, the research

Table 1 Module inflation information analysis
Table 2 Module internal pressure information analysis work presented a scheme of technological realization and a systematic approach for deformed architecture using cybernetic robotics.In this study, the testing was extended to diverse surface variations among the materials, and open-ended possibilities for a series of unexpected scenarios to create free-style and abundant shapes of space by developing the combination of multiple samples.The mode and level of aerodynamics affect the shape and scale of the space, which will correspond to different scenes.Under the complete mechanism, the data of human activities captured by intelligent perceiving sensors are linked layer by layer to the space they need or will need and its behavior is triggered by sensory feedback, forming a completely closed loop of spatial interaction sustainability.Based on this approach, the prediction of the performance of the model allows indeterminacy, and the equilibrium of the improvisation in the activity and scene adaptability through the synergistic behaviors of these samples is obtained.Besides, the time-sharing, interactive computer regime is introduced for the dynamic in the equilibrium.These findings give a resonant envision of the future of intelligence in architectural environments and extend pneumatic robotics architecture, confirming a more flexible, circumstantial, and biotic facet of architecture as an interactive environment.
Although the hypotheses of the SPRS were supported statistically, indeterminacy, the crux collectively pointed out by this work and the relevant research seems contradictory to the cybernetics mechanism.As Astrid mentioned, while material agency denotes the possibility that things can act, material objects have an effect on the course of action that is irreducible to direct human intervention (Oyen, 2018).Thus, the counterbalance between the morphological manipulation of the cybernetic system and the autonomy and randomness of space usage calls for discreet consideration in human-system interaction.Future work should include the following focus points.The research on the fabrication and performance of the soft modules on the architectural scale is to be furthered; What to be mentioned gladly is, as the developments require, the emergence of several integrated development platforms for soft robotics, such as FlowIO, enabling users to effortlessly bring to life nearly any project in the soft robotics and programmable materials domains (Shtarbanov, 2021).The specific performance strategy of the space interaction, which relied on cybernetics as a dynamic system including behavioral goals out of realistic flow standards, should be explored.Furthermore, oriented on equilibrium and inclusivity, a resilient control system based on feedback investigation, game theory, and behavioral science can be introduced to study soft architecture with long-term performative strategies.
What's notable is the study investigates the soft robotics architectural system by integrating multi-faceted impact upon intelligent architecture in material, space, interaction, fabrication, and cybernetics.As the structure with its inhabitants and the interplay between them form the entity of an evolving ecosystem mediated by each of them, the idea of "system evolves" will lead the architecting process beyond architecture.Architecture acted as the pioneer for the vision of augmenting human intelligence because of its power of building worlds.Therefore, architects need to reconsider the scope of their work and the responsibility for environments in the world shaped by computer programs increasingly and indirectly.
Architecting is concerned with human use, element organization, construction, and also conceiving, designing, and interpretation.Architects are more trained to build the order of space for its interpretative performance throughout time, such as setting the order of four seasons, and nature will generate landscapes itself.Just as an algorithm allows the program to generate outputs with inputs, code is defined as the space for its capability to shape the environment.In this way, the physical information finds a way to correspond with space performance through architecture algorithms.The SPRS project is based on the transcending potential of the usability of artificial intelligence and cybernetics have for architecture.
SPRS provides examples and observations in different dimensions of architectural intelligence extension through soft robotics applied in architecture.It sets inputs that connect initial artificial intelligence, especially the intelligent brain.The basic setting of the system is body-tracking and gesture recognition, and further design can be extended to visual capturing, heuristics, machine cognition, and communication systems.The SPRS configured outputs of primary robotics and pneumatic systems related to cybernetics, which acts involving information architecture and spatial research, and its ideal practice is framed on the feedback mechanism and real-time interaction system.SPRS applies human-computer interaction to architecture and integrates multi-dimensional perception, multi-variable analysis, and multi-level command into its responsive intelligence.In addition, it can extend its practicing connotation in an intelligent environment convergent of spatial and informational interfaces that city and architecture are to form as the carrier of information and space memories.In this situation, SPRS switches to different scenes to accommodate various and multivariable life-conditioning by issuing commands and gestures to the system.The convergence of spatial information makes everything interconnected, and the shaping of future communities and urban landscapes with it will change the ecology of human settlements.

Conclusion
Based on the demand for flexible and dynamic living space, this research aims to construct a changeable architecture based on soft robots.It is sensitive to the environment and equipped with responsiveness, plasticity, and transient structure through innovative ideas and technologies, designated "responsive architecture" or "soft architecture".In this experiment, from the human data perception and feedback analysis to the responsive mechanisms and the generation of spatial data, the material dynamic closed-loop process of humans and space is digitally synchronized and created, laying a technical logic foundation for the intelligent architectural system.The construction of the system model starts from the structure and performance, linking data of the human body and each physical level of the space, thus forming an interactive information base and a complete agent chain.The responsiveness of the system requires the perception and interpretation of human activities and their demands, feedback action mechanisms timely trigger with corresponding changes in space, and artificial intelligence algorithm settings based on high-dimensional definitions towards the scene presentation.Although the combination of the perception system, data links, pneumatic device, and space scene requires extensive cooperation and deeper exploration from more artistic and technical strength, the model framework established by this design is rather innovative and effective in the long run.
As information is interconnected and assembled in this age, the interface between people and the environment tends to be multi-layered and invisible, and the entity of the building radiates the surroundings in a more flexible and potential way.The soft robotic architectural system is designed to simulate and prepare for the space required for human dynamic living as much as it could, it is hoped to expand the possible future of human settlements of soft architecture through soft robots and further explore the integration methods and connotations it has been and can be incorporated along the way.

Fig. 3
Fig. 3 Cedric Price, Fun Palace, sketches and notes, 1964, Cedric Price Archives, Canadian Centre for Architecture, Montreal

Fig. 4
Fig. 4 The performance change under the same structure in different time zones indicates the form's competence and its different interpretation

Fig. 6
Fig. 6 Shi, 2018, Robot No. 6 "Soft: Balancing and Wiggling" The kinetic actuation can be collected from the scene switching (e. g. reshaped lounge/partition rise up/seating zones of different undulating surfaces), and the output devices include an actuation motor, kinetic driver, VR glasses, odor transmitter, temperature regulator, glass transparency regulator, tilt brush, music player, visual projections, and view wander.The interaction input can be set by fingerprints, VR waves, body temperature, sunlight, voice, panel setting, press projected spot, facial capturing, and body tracking.An interactive performance can be better actuated and set into different levels by synergizing all these mentioned points.It is expressed as follows: Level 1-Interaction in the unconscious state: Pressure controlling (spontaneous deformation): Sofas and beds for sitting and lying; Level 2-Interaction in transforming mode: Body-pose controlling via skeleton tracking (smart deformation): Space transforms into different scenes.

Fig. 15
Fig. 15 Sample swelling testing and morphology detection

Fig. 16
Fig. 16 Space generation and scenario presentations