Collaborative Research: Merging Human Creativity with Computational Intelligence for the Design of Next Generation Responsive Architecture

该项目将资助一项研究，该研究将人类设计师的创造力和经验与计算智能相结合，以提高环保建筑技术设计的新奇和有效性。 建筑运营占基础设施能源足迹的很大一部分。 自适应建筑组件，如折纸灵感的自适应立面，是降低这一成本的潜在途径，通过使建筑外部能够根据光或热条件进行重新配置。 然而，阻碍这些概念的一个主要挑战是它们无法被系统地设计。 此外，迫切需要提供早期和互动的设计反馈，以更好地促进更有用，细致入微和生态响应的想法。 这项研究将创建促进创造力，知识共享和集体工作流程的工具，旨在从美学到功能的各个方面实现有效的技术。 这项工作将同时为建筑和工程专业的学生创建一个新的教育轨道，以了解和测试他们对自适应建筑概念的想法。 通过提供测试平台和关键数据集来告知计算工具，这种教育工作完全融入了研究。 因此，这项工作将为折纸启发的结构创造新的设计过程，并为未来几代设计师提供探索自适应结构概念的技能和爱好。 这项工作的总体目标是建立一种新的策略，将建筑设计过程和人类衍生的设计与物理指导的机器学习和计算设计优化相结合，并评估这种策略在折纸启发的响应式建筑中的应用。计算-实验一体化的研究工作包括三个方面：1）建立一个人机耦合的设计策略，使折纸自适应结构能够同时满足多个外部环境交互目标和内在结构要求：2）建立一个计算框架来评估折纸自适应结构的性能，并基于人的输入和计算机的指导来增强其设计;以及3）建立一个在建筑设计过程中引入新的结构概念的框架，并应用该框架来识别适应性结构的新兴折纸灵感概念。 该项目旨在建立一个总体框架，通过计算工具的新颖组合来加速和扩大创新，这些工具不仅与用户相连，而且还可以增强人类的创造力。除了创建新的计算工具集外，这项工作还将生成独特的数据集，以利用一系列机器学习工具，并发现和验证具有独特优势的新折叠模式。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This project will fund research investigating creation of a process combining the creativity and experience of human designers with computational intelligence to increase the novelty and effectiveness of designs for environmentally responsive building technologies. Building operations account for a large portion of infrastructure energy footprint. Adaptive building components, such as origami-inspired adaptive facades, are a potential pathway to reduce this cost, by enabling building exteriors to be reconfigured based on light or thermal conditions. Yet, a major challenge holding these concepts back is their inability to be systematically designed. Moreover, there is a critical need to provide early and interactive design feedback to better facilitate more useful, nuanced, and ecologically responsive ideas. This research will create tools that promote creativity, knowledge sharing, and collective workflows that intend to lead to effective technology in all regards, from aesthetics to function. The work will simultaneously create a new educational track for architecture and engineering students to learn about and test their ideas for adaptive building concepts. This educational effort is fully integrated into the research by providing a testbed and the critical dataset to inform the computational tools. Thus, this work will create new design processes for origami-inspired structures and seed future generations of designers with the skills and penchant to explore adaptive structure concepts. The overall goal of this work is to establish a new strategy to integrate architectural design processes and human-derived designs with physics-guided machine learning and computational design optimization, and to evaluate this strategy on the application of origami-inspired responsive architecture. The integrated computational-experimental research effort includes three thrusts: 1) establish a coupled human-computer design strategy for component-level origami-inspired adaptive structures to simultaneously meet multiple external environmental interaction goals and intrinsic structural requirements; 2) create a computational framework to evaluate the performance of origami-inspired adaptive structures, and augment their design based on human input and computer guidance; and 3) establish a framework for the introduction of new structural concepts in an architectural design process and application of this framework to identify emergent origami-inspired concepts for adaptive structures. The project intends to establish an overarching framework that accelerates and widens innovation through a novel combination of computational tools, which not only connect to the user, but can also augment human creativity. In addition to the creation of new computational toolsets, this work will generate unique datasets for leveraging a spectrum of machine learning tools and discover and validate new folding patterns with uniquely beneficial performance.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.