As layers of networks and information system genes are currently emerging in the design of spaces, new approaches to the study of architecture are emerging too. The way we describe and understand buildings is being radically transformed in the Internet of Things (IoT) and Internet of Places (IoP) era. Understanding how data is affecting the making of design is important to building “intelligent” buildings. The current dilemma in smart cities and buildings initiatives has been caused by data-IT-enabled processes, through app design and data visualization, rather than design enabled processes. Design through performance rules, if coupled with information design modelling and Machine Learning mechanism, can allow real “smart” buildings and cities. Through academic and practice case studies, our research aims at drawing a framework of teaching and designing real-time data-enabled building envelopes which are controlled by data inputs. It will project insights on the theory of cybernetic and adapting architecture where geometry and building components can change to allow better performance in real-time. With a focus on building envelops, this paper discusses the challenges of end user and post-occupancy real-time architecture.
Adaptability, flexibility, and performance have played an important role in experimental architectural projects over the last decade, resulting in a shift towards kinetic, rather than static, environments . Whilst biomorphism is a source of design inspiration, we have seen, since the 1960s, a burgeoning of visionary projects influenced by nature that focus on form finding in the built environment. Investigating the overlaps between biology and architecture, we find that a biological paradigm inspires the current frontier of research and innovation in many sectors. Meanwhile, the architectural landscape has adopted nature inspired designs as valid strategies. However, it still lacks a showcase of innovative products or a real breakthrough in the form of a real, naturally performative or adaptive architecture. For the most part, the applied research developed in the name of adaptive architecture and kinetic facades lacks the dynamics which would allow smart architecture to possess the ability of auto-initiative and self-learning, fulfilling the prophetic claim that “the ultimate smart structure would design itself”. The paper reviews recent experimental projects which explore the potential of using adaptive real-time architecture envelops, it discusses the recent development of Machine Learning in designing real-time architecture spaces. The paper discusses the challenges that the built environment is facing in order to implement such an approach.
Framework to connect the making of architecture with smart buildings
Machine Learning provides computers with the ability to learn without being explicitly programmed. Machine learning focuses on the development of computer programs that can teach themselves to grow and change when exposed to new data. The integration of data-driven geometry that has the capability of changing its behavior and adapting to environmental changes to provide a better performance of the building is needed. New concepts of smart materials and building components for building facades will be explored in this paper by analyzing the grammar of each building component during the design process and the application of smart material and the feedback loop during building operation through selected case studies.
The ability to control and/or understand data through supervised machine learning algorithms, where we have an assumption about the relationship between the input and the output, will introduce new generations of CAD software that is based on ongoing feedback loops that aim to refine design through the design phase by harnessing many factors that affect building performance and also through the operation phase. This would challenge the current practices of architecture and would offer different modules of design and post-occupancy automation processes.
Our research explores Ambient Intelligence (AmI), a full integration of technology and knowledge opens the possibility for collaborative, and the integration of such process in the making of architecture envelops in order to be able to take over tasks when other systems drop out, and anticipate the desires of the occupants and feedback from the environment with a conscious mediation.
In our applied research we aim to provide a theoretical framework by making a connection between real-time data-driven building envelops and post-occupancy automation, the paper will describe the theory of cybernetics and it will build on the recent development of the work in tangible computing which will allow end users to decide the current state of the art of their building design. The suggested framework of adaptive architecture spaces would evolve into what Catherine Malabou in her book, What Should We Do with Our Brain?, explained as plasticity; “the quality by which our brains develop and change throughout the course of our lives.” ‘Intelligent’ and real-time architecture would be able to not only adapt to existing circumstances and user desires, but would possess “a margin of freedom to intervene, to change those very circumstances” and create an understanding that would open-up “a newly transformative aspect of the neurosciences.”