YEAR:                      2015

TYPOLOGY:          Research project

TEAM:                   Matteo Silverio, Josep Alcover,                                    Ji Won Jun

ACHIEVEMENTS:  Exposed at Construmat, BCN

                                 Innovative Building Award

The project implements shape memory alloys, digital content, fabrication and user interfaces to set up an intelligent responsive structural system that can be applied in different architectural scales. 

Remembrane is an exploration of applying shape memory alloys to create a big scale adaptive kinetic pantograph-tensegrity structure that could adapt to environmental changes and be easily controlled by users through a user-friendly interface. The research focus has been placed on the structural design following the material properties and allowing users to be the ultimate decision makers of the performance of the structure. The developed prototype investigates the use of springs made of Nitinol as linear actuators. The project avoided from the beginning the use of centralized heavy motors to move the system and focused on developing a distributed system of lightweight actuators (nitinol springs) embedded into the structure that allow a much more precise control of the shape and a more energy efficient movement. 

The interaction with the structure and the different ways of controlling it have also been an essential aspect of the project. Easy manual control of the shape and automatic autonomous performance of the system had to be combined in order to achieve a truly intelligent and efficient interaction.  Both ideas were at the starting point of the design of the user interface.  The final prototype of the research is a real scale performative kinetic structure. The prototype has been deliberately left without a specific function in order to allow us to envision an endless number of potential applications: from adaptive partition solutions to environmentally responsive roofs, to morphing facades and changing urban elements.


The first development step was to test different actuators and assess them in terms of their resistance, lifespan and lightweightness. Shape memory alloys (in particular nitinol) were the selected solution because they are relatively powerful and lightweight. These two characteristics make them very suitable for an actuation system that is embedded into the structure and distributed throughout its different parts. Nitinol wires are flexible at normal air temperature and they go back to a “saved shape” when heated above the transformation temperature (70 ºC for the used wires). This transformation process can be used to create an actuator and enhanced by shaping the wires into springs. 
Several experiments were carried out to test the material resistance and strength in order to optimize the structure and its movement. Moreover, a proper heating system had to be developed to optimize the reaction time as well as to avoid material damages, trying to make the solution reliable and stable. The tests proved the material capacity to lift certain weight when heated.
The results underline a relationship of direct proportionality between spring’s shrink and loads. However, tests have highlighted how unstable the material is if not properly heated. Therefore, to avoid any material damages, a proper and controlled heating system had to be designed. Activating Nitinol using DC current with the proper amount of electricity allows a homogeneous and precise heating. However, DC current could easily overheat and damage the Nitinol wire. To limit these potential material damages a Pulse Width Modulation (PWM) circuit was developed. 
Using PWM has several advantages: it turns the current on and off to the wire very quickly and this power oscillation allows to keep heating the wire while avoiding damage to the material’s crystalline structure. Moreover, by modulating the “oscillation” period, the shrink time can be easily controlled. 
After several tests and physical models, springs made of 1 mm thick Nitinol were used for the prototypes because they presented appropriate strength for the scale that was being studied. The electronic circuit designed to control these Nitinol wires was composed by an Arduino board and four transistors that controlled the amount of current released in each Nitinol wire. 


After studying the properties of Nitinol a series of tests focused on geometric principles and structural systems that allow kinetic behavior have been carried out. 

The geometrical system that presented the best performance was a diagrid pantograph with pseudo-tensegrity principles. Wood pieces with flexible plastic joints and a diagonal network of Nitinol springs were the main components of the first prototype (Remembrane 1.0). 
The bending of the surface is provoked by the contraction of the Nitinol springs when they are heated. This first prototype was a significant step in the study, however, there were four main aspects that had to be improved or developed: improving the actuation and creating a locking system, reducing the weight of the structure, designing and building a skin and programming a user interface.
The structural optimization has been an essential step to maximize the lightweight performance of the structure. In order to minimize the system weight without modifying its strength, a topological optimization process has been carried out. This analysis has generated an important mass reduction (about 52%). 

To perform the structural analysis, Karamba, - a parametric structural engineering plugin for Grasshopper for Rhinoceros3D - has been used. Thanks to this tool, the structure has been checked from a structural point of view, and the shape of each element has been informed according to the software results, minimizing the material quantity and improving the general system behavior.

By reducing the structure weight, Nitinol springs need less force (and electricity) to move the structure. Moreover, each joint has been informed by the analysis optimizing its shape and minimizing deformations.


A truly innovative responsive kinetic system has to be easily controllable but also intelligent enough to make its own decisions. In order to achieve that, a complex interaction system had to be designed. The system has two possible ways of interacting: interface mode (the user sends orders through the user interface) and sensing mode (the structure reacts to the sensors’ inputs).
To make this complex interaction as easy as possible a user friendly web based interface has been designed. A web application has the huge advantage of being accessible by any device that can support a web browser. This means that any computer, smartphone or tablet can be used to run the user interface and no special software or hardware is needed. The interface has been developed in order to not only send orders to the prototype but also to visualize the information measured by the sensors installed on the structure. 
To physically interact with the structure a microcontroller was needed. The microcontroller acts as an intermediate between the web application and the prototype. The most extended and easy to use microcontroller is the Arduino and among all the Arduino models the Yun was chosen because it can connect to the internet using wifi and, therefore, it allows to interact with the web application wirelessly. The actual interaction with the Remembrane takes place at the input and output pins of the microcontroller. The input signals come from 2 proximity sensors (one in each side of the structure) and 1 accelerometer that measure the inclination. The output signals go to the network of actuators, to the locking / unlocking system and to the 2 LEDs that show the status of the structure.
The GUI has been carefully designed to allow an easy and efficient interaction between the user and the Remembrane. It is based on interactive graphics that make the whole communication process very intuitive. The most important element of the interface is the interactive 2D diagram. It can be reshaped just by touching the screen of a tablet/smartphone (or clicking and dragging on a computer) and it automatically sends the necessary information to shape the structure in the same way. The 2D diagram is complemented by 2 other elements: a visualization of the real inclination of the structure and a visualization of the distance to obstacles. The 2D diagram is, therefore, a tool to send instructions but also a tool to understand the current state of the prototype. 
On the top left corner of the interface there is a button that allows the user to switch between interface control mode and sensing mode. In the interface mode the structure reacts only to the instructions sent through the interface. On the other hand, in the sensing mode, the prototype performs autonomously according to the sensors’ inputs. In fact, it behaves according to a set of rules that have been pre-programmed. The final prototype has a very basic set of rules: when the proximity sensor detects that something is very close, the structure bends to the opposite side. This simple behavior has the purpose of illustrating some kind of autonomous behaviour that could be further developed.



Within the framework of the structural investigation, alternative systems have been explored. 
The aim of this research part has been the study of different and freer structures able to achieve double curvature as well as a biaxial deployability.

The developed structure has two main type of actuators: one allows the rotation between elements. The other one allows the system deployability independently along x and y. As the previous ones, the new structure is lightweight and characterized by a distributed system of sensor and actuators that permit to achieve any kind of form.

The vision of this project is beyond the idea of intelligent façades or similar systens.  Our vision is an intelligent membraneous structure that redefines space and architecture that adapts to human demands and environmental impacts. 
We would like to see an evolution towards a society with interactive, responsive and spontaneous architecture, where users participate in the creation of their own customized and personalized environment. 

Furthermore, the living space becomes a living organism which has its own intelligence and optimally perform with least energy, continuously adapting and regulating its inhabitants and their comfort. 

The exponential growth of technologies will then bring humanity to a point, potentially in near future, where the built environment is in harmonic modulation with its occupants’ desires, where it has acquired an artificial intelligence, it shall interact and respond to human thoughts through brain-machine interface will be the singularity where the boundary between human and machine will transcends, where our living habitat will be one with our thoughts, even greater than our intelligence. 
Beyond the pragmatic reasons and practical pursuits, a new relationship with our built environment will be part of life, evocatively and emotively with enhanced experience, memory and senses.