Cyberfafacade —— Light Sandbox (Shuangying Xu, Yuxin Qiu, Sienna Xin Sun)

Facade, as part of the building envelop, is the shelter for us from the noisy world outside as well as the annoying weather conditions. However, a solid facade will lead to both physical and social isolation. Research shows that isolation and loneliness will cause health problems, including depression, sleep disturbances and decreased attention and learning ability.[1]. Daylight is thus acturally a basic necessity of life.

The pandemic has largely changed the way of living and working, when nowadays more and more people are choosing to workat home. Therefore, an effective facade, with the functions of constantly adjusting and coordinating the incoming daylight for different daily activities, is necessary.

In this project, we take sunlight as a lighting resource to be manipulated by the facade and used by people. The aim of this project is to create an adaptive facade with the function of manipulating sunlight, view-friendly and vivid in form. In detail, the facade should be able to achieve real-time, continuous adjusting of light conditions automatically and moderately. The adjustment should be flexible and targeted enough when there are different requirements in different areas at the same time. Furthermore, to create a user-friendly facade system, it should be light weighted, unit-based, and easy to assemble and maintain. Here we take modular inflatable building components as a good solution.

In the „Sunlight Sandbox“, three different modes are preset as shading, moderate, and transparent and will be achieved as a result of the interaction of sun conditions and light requirements. The double-layered inflatable modules will be functionally inflated and deflated, where the inner non-transparent coloured layer will create adaptive shading reacting to human activities as well as a vivid tactile atmosphere for the residents inside.

In order to get better control of the system, we applied here Agent Based Modeling and Simulation(ABMS). ABMS can be seen as a variation of particle-based modeling approaches, which can be used to model and simulate complex physical systems, but where the constituent entities (particles) are endowed with decision-making capacity and the ability to adapt their behaviors. In our case, the continuous changing light conditions and human requirements will keep interacting with each other. By using ABMS, we let the inflatable units make decisions by themselves and adapt their behaviors according to both input changes, which is able to ignore most of the inherent sequences of interaction and thus solve the problem of differential updating frequency of the inputs.

We take each unit of the facade surface as a boid agent. For every iteration, it will be influenced by inputs of sunlight conditions and human requirements detected by the sensors. Each sensor will have an influencing area of 16 neighboring agents, which consists of a single control group. Within each control group, there will be a hierarchical effect on each agent in terms of the distance between the sensor and agent positions, which will lead to a sequencing inflation or deflation action. Among the control groups, the values of sensors in the neighboring group will also influence each other with different weights according to the distance in order to achieve a continuous and gentle change on the facade. Additionally, it will also minimize the negative effects due to the unstable or unreasonable sensor feedback.

Possible future development can be to add a feedback port for the users to evaluate if the facade system has fulfilled their requirements or not. This will hopefully help to realize a learning capability for the intelligent facade system.

The result of the ABMS system will be a boolean signal of inflating or deflating for each boid agent in a mesh environment, which has a basic behavior of containment that helps to regulate the maximum shape change of each inflatable unit.

In this project, we use light sensors on the outer side of the facade to get real time light intensity values as sun conditions. To get information about human requirements, we use webcam color detection to recognize the location and the motion of the users.

To get the facade units to inflate, vacuum pumps are introduced together with the solenoid valves for deflation. All of them are connected to the relay module and then to the microcontroller.The sensors and actuators are controlled through Grasshopper plug-in Firefly, in which the sensor values are sent to the ABMS System and then give out the inflating and deflating decisions to the actuators.

The whole facade consists of standardized double layer inflatable modular units. For each unit, a branching pipe will go through the outer layer to realize the inflation or deflation of the inner layer. The space between the two layers is treated as most possibly vacuum. The branching pipe from each unit will then go along with the supporting structure and flow into the main pipe of the respective controlling group, which is connected with the pump for inflation and valve for deflation.

The modular inflatable system will be installed together with a post and beam facade structure, on which the sensors are fixed and along which the cables are tracing. The whole structure is self-standing and lightweight, which will be easy to be prefabricated and assembled. Each unit will be replaceable and friendly for maintainence. In addition, the color of the inner layer can also be customized according to the personal preference.

The rendering simulation takes place based on real life scenarios. Taking a family of four as an example, here shows a typical summer day.

_08:00 Read / Wander

 Moderate light

_09:00 Read / Play / Work

  Light needed

_12:00 Play / Work  

    Light / Shading needed

In the afternoon, due to the strong sunlight, the performance of the facade will be more likely to provide enough shading for the user inside.

_13:00 Work / Play 

  Shading needed

_16:00 Rest / Handwork

    Shading / Light needed

_18:00 Family Activities

  Sunset

For physical prototyping, we take a set of eight units controlled by two control groups to show the sequencing within and interaction between different groups. 

In this project, we use clear vinyl sheets as the outer layer of the units, while colorful balloons are used in the place of expensive colored silicon foils as the inner layer. The pumps and valves are connected through silicon pipes and 3D printed connectors used for diverting the air flow to or from the balloons inside the units 

Through the tests we have found that the length and diameter of the pipes had significant influence on the sequencing of inflating and deflating.

To simplify the process of the physical tests, both of the light sensors and the webcam are attach to the same side of the prototype temporarily. 

The first physical test shows a series of reaction from the facade when a user is passing by.

The balloons have an initial state of semi-inflated due to the moderate light condition. When the user go towards the facade, in order to provide more light for the human activities, the balloons inside start to deflate. When the user leave the detectable area, the facade will go back to the initial state.

The second physical test shows the case of a different light condition with the same user activity.

In this case, the initial state of the balloons are fully inflated due to the stong light detected by the light sensor. When the user pass by, the facade also intend to adjust the amount of incoming light. Therefore, the deflation of the balloons are obviously slighter than that in the former test.