reTessellate: Modular Dynamic Surfaces Reactive to Socio-Environmental Conditions

Abstract

This paper presents the design, fabrication, and programming of a reconfigurable structure. The structure is comprised of modular, programable units, which can be stacked together, and collectively construct surfaces that shape-shifts in response to the surrounding physical and social conditions aiming at enhancing the link between occupants and the built environment.

1 Background

Reconfigurable structures are well established within our built environments and exist all around us. For instance, doors and windows are some of the earliest examples of reconfigurable structures that provide a dynamic manipulation of the building layout to satisfy various conditions. Examples across different building typologies exist from tiny houses (in which reconfigurable structures play a key role in space planning) to public spaces. In airports, for instance, passenger boarding bridge changes shape and position to facilitate a convenient and safe connection between the building and the aircraft (Fig. 1).

Fig. 1.

Passenger boarding bridges (image credits: Wikimedia).

With the technology being increasingly embedded in various aspects of our lives, reconfigurable structures have become more responsive and intelligent in our everyday life. For instance, doors can sense an incoming motion and react accordingly. Building envelopes can dynamically control the amount of light that enters the building. Additionally, mechanical systems can be added to such list since they allow for adaptation of interior quality in response to climatic conditions and human behavior. Building components such as windows, walls and movable partitions, doors, screen systems, louver and shading devices, and operable vents can allow for adaptation of the building in response to human needs.

The integration of responsive features in Architecture, offers the opportunity to experience buildings as living, changeable organs versus buildings being static entities. Buildings with reconfigurable structures can be designed in a way that their appearance, permeability, and affordances can (un)predictably change in response to environmental changes and their occupants. Le Corbusier, famous Swiss-French architect and the pioneer of modernist architecture (1887–1965), used the term “household equipment” and referred to the house as a “machine for living in” to define an architecture that involves operable and reconfigurable elements. In the 1970s, avantgarde thinkers like Nicholas Negroponte suggested that advancements in computation and artificial intelligence would soon make buildings smarter while being capable of intelligently recognizing users’ activities and environmental changes for a more customizable and adaptive living condition [1]. “Negroponte proposes that responsive architecture is the natural product of the integration of computing power into built spaces and structures. He also extends this belief to include the concepts of recognition, intention, contextual variation, and meaning into computed responses and their successful and ubiquitous integration into architecture” [1]. Today, automated reconfigurable structures have become more commonplace and are frequently integrated in building envelopes as a tool for enhancing human comfort and energy efficiency.

In this paper, what we refer to as reconfigurable and responsive structures encompasses all building elements that are able to change and adapt in real-time in response to socio-environmental conditions in order for accommodating the contingencies of everyday life. We argue that computation and reconfigurable structures can converge and be embedded at different scales of the built environment, in particular, architectural surfaces, which can produce more flexible, responsive, and intelligent spaces aiming to optimize our social and physical experience of architectural spaces that we live in.

2 Research Through Design Approach

The field of Human Computer Interaction (HCI) is experiencing an increasing interest in Research through Design (RtD) as a research method which can offer distinct advantages to the HCI community. RtD’s contribution to the HCI field has been characterized in different ways and many of them view RtD as a canonical type of design-research activity. For instance, Frayling [2] referred to developmental work which involves “a customization of a piece of technology to do something no one had considered before, and communicating the results.” In action research, “a research diary tells, in a step-by-step way, of a practical experiment in the studios, and the resulting report aims to contextualize it to communicate the results” [2]. In general, RtD offers researchers the opportunity to cope with messy design-research problems that are unclear or not well situated in other methods of research. RtD also enforces designers and researchers to focus on the work of future, instead of the present or the past. Additionally, RtD makes designers think about the logics and ethics of the design process rather than solely focusing on the final outcome [3].

This paper presents an exemplar case of research through design as a result of an iterative design process. The proposed, fully functioning prototype was designed and tested during different phases as we improved the usability and efficiency of the system.

3 The Backstory

The design-research work presented in this paper started as a response to social conditions in public spaces aiming at answering the question of “how can the environment adapt to the human needs within public spaces?” The theoretical stance builds upon Christopher Alexander’s concept of positive and negative spaces. In his discussion of positive and negative spaces, Alexander argues that “people feel comfortable in spaces which are positive; and people feel relatively uncomfortable in spaces which are negative”. He later established a link between the positive-negative concept and the shape of the indoor walls in which he explains that straight walls “make no sense in human or structural terms” and to create a social, inviting space the shape of the wall needs to be concave when possible (especially in thick walls) [4]. The Reconfigurable Wall System (RWS) came as a response in order to provide various configurations offering both feelings interchangeably (Fig. 2). The configurations are (1) embrace, (2) repellent, and (3) delineate [5].

Fig. 2.

The Reconfigurable Wall System (RWS). Top right: conceptual design; top left: programing and sensing experimentation; bottom: fully-functioning prototype.

This paper revisits the RWS and further elaborates the opportunities to design a flexible, intelligent structure, which can be integrated in various parts of the built environment while responding to various socio-environmental conditions.

4 Tessellations in Nature as a Source of Inspiration

Nature provides a fascinating source of inspiration for pattern development, surface subdivisions, and close-packing, i.e., tessellations. The existence of tessellations in nature relates, in many cases, to optimization purposes by producing “economical structural solutions to some given set of conditions” [6]. Various examples exist in nature which have been shaped, evolved, and refined over time through evolutionary processes. Some tessellations can be seen directly with the naked eye such as the honeycomb structures, patterns in animals’ skins, and cracked mud; whereas, some are invisible to the naked eye, such as the living cell structures [6] (Fig. 3).

Fig. 3.

Tessellations in nature. Top left: hexagonal pattern in a honeycomb; top right: microscopical leaf structure; bottom left: cracked mud; and bottom right: Voronoi structure on giraffe’s skin (image credits: Wikimedia)

This paper employs tessellations to subdivide surfaces and focuses on three types of tessellations: regular tessellations, semi-regular tessellations, and tessellations with regular polygons. Mathematically, tessellation is defined as the repetition of polygonal shapes infinitely to fill a surface area while leaving no gaps between the polygons [6]. For regular tessellations the surface is subdivided into smaller sections using identical, congruent regular polygons. There are only three possible scenarios in the case of regular tessellations: triangular-, square-, and hexagonal-based subdivisions. In semi-regular tessellations, more than one congruent regular polygon is employed. However, the collection of geometries within the subdivided surface should be identical from any point in the surface, otherwise the type of tessellation will fall under the category of tessellation with regular polygons [6]. Both semi-regular tessellations and tessellation with regular polygons provide a good opportunity for morphing different tessellations within a surface area (Fig. 4).

Fig. 4.

Regular tessellations and semi-regular tessellations.

5 Design Process

Our team investigated the possibility of extending applications of reconfigurable structures in different building components. Through investigating various tessellating systems, this paper presents a Dynamic Modular Tessellation (DMT) system–reTessellate—in the form of units that can be stacked to form a structure, which can be used in various architectural surfaces e.g. celling, building façade, and walls.

Significant hardware and software modifications were conducted on the RWS. The following sections discuss detailed modifications.

5.1 Hardware Design

In order to create a flexible surface structure which fits in a variety of situations, the design of the new system hardware has to be flexible enough to cover various type of tessellation systems while prioritizing “minimum inventory/maximum diversity” design strategies.

The current system hardware consists of three main components: a repetitive unit, a repetitive structure, and the core units. The repetitive unit is the basic building block of the system that hosts the core unit (Fig. 5). The repetitive unit can be disassembled into two pieces—male and female—when necessary. A consideration was given to how this geometry can be repeated to construct various novel tessellations.

Fig. 5.

Top: the repeated unit; and bottom: the core and its installation.

The repetitive structure is the product of continuously stacking the repetitive unit to fill the required area. The system can be assembled to form all the possible regular tessellation structures (triangular-, square-, and hexagonal-based tessellations), semi-regular tessellations, and tessellations with regular polygons. This feature significantly increases structural flexibility and provide an opportunity to seamlessly morph between various types of regular tessellation systems (Fig. 6).

Fig. 6.

Top: assembly of the triangular-based repetitive structure. Bottom: the first row, employing the DMT units to assemble triangular-, hexagonal-, and square-based tessellations; the second row, morphing the structural system.

The core, on the other hand, is also modular, and different modules can be attached to the repetitive units to offer a variety of reconfigurable and transfigurable environmental settings. For instance, the module presented in Fig. 7 (Left) is a linear actuator module. This module function in a similar way that the RWS works (a push and pull mechanism). The study also examined the design of other modules that can control origami folding star units as a responsive shading device (Fig. 7).

Fig. 7.

Prototyping and testing of the DMT unit. Left: linear actuator module. Right: origami-star as a shading device.

5.2 Software Development

The Dynamic Modular tessellation system was programmed using Rhinoceros’ Grasshopper and Firefly. The latter acted as the primary communication tool between Grasshopper and the Arduino microcontroller. Two algorithms were developed to provide two primary dialogue modes:

(1) the preset interaction; and (2) the responsive interaction (sensing the environment).

In the preset interaction mode, the structure shapeshifts based on hardcoded values. This mode provides testing environment of the system functionality including both hardware and software (Fig. 8).

Fig. 8.

Preset interaction mode (preprogrammed wave).

In the responsive interaction mode, the system reconfigures in response to the data collected from the surrounding environment through embedded sensors. The structure integrates a Microsoft Kinect motion sensor, which is utilized with “skeletal mapping” and depth information within the Grasshopper plugin to understand nearby occupants (Fig. 9).

Fig. 9.

Responsive interaction mode

6 Discussion of the System’s Current State

The DMT system offers two types of configurations: per-configuration and post-configuration. Both configuration modes depend on expected activities within the space. The pre-configuration mode is the initial setup of the system. In the pre-configuration mode, the designer gets to choose the type of tessellations and the installation modules. The post-configuration mode is the capability of the system to shapeshift to accommodate post-occupancy needs.

Tessellations offered many advantages to the design of the DMT system. A surface that is constructed from the DMT units provides an excellent opportunity to be reconfigured and customized during the design and fabrication phase using “minimum inventory” of the modular units. The DMT surfaces is capable of morphing the internal structural subdivisions of the surface between various stacking systems. Such a feature makes it possible to have a single responsive surface which offers different configurations at different points while being controlled by a single control center. For instance, consider the corridor presented in Fig. 10 which reimagines passengers’ corridors public spaces. The DMT system is forming a surface that is being morphed from square-based tessellation to a hexagonal based tessellation (i.e., aperture system). The wall structure aims to better accommodate passengers’ activities in crowded corridors while the aperture system controls light intensity in the interior spaces.). The aperture is also capable of morphing into a triangular-based tessellation.

Fig. 10.

Rendering of the dynamic modular tessellation (DMT).

7 Future Works

The next step is to conduct usability testing and additional software adjustments. The current modular design provides a great environment for testing the system in different scenarios. We aim to enhance the efficacy of the DMT System by conducting different usability studies. User groups will be recruited to evaluate the system usability and functionality.

Additionally, the DMT system is envisioned to be able to teach itself and improve its behavior over time. Our next step also includes the development of “algorithms that improve through experience,” which ultimately contribute to a more intelligent dialogue between the occupants, the structure, and the built environment [7].