Application of Design Data Analytics in Computational Design of Built Environments

设计师探索多种设计方案。他们从不完整、不精确和不断变化的设计目标开始，将复杂的新兴设计数据提炼成有意义和实用的设计理念，以找到“令人满意”的解决方案。在这里，我们使用“设计数据”指的是外部代理提供的在设计决策中考虑的任何输入，或在过程中生成的任何输出数据，并作为输入提供给设计。当前的设计系统缺乏对基于数据的设计探索的支持：首先，它们只支持单一状态模型，迫使设计师按顺序工作。其次，它们缺乏处理大量、多样、动态的设计数据的功能，这些设计数据随着替代方案的探索而迅速增长。这些缺点是相互关联的，必须通过改进系统来全面解决，以更好地支持使用替代方案。因此，本提案提出了设计新的计算设计工具和工作流程以处理替代方案的需求。当大量的替代方案成为设计探索的组成部分时，这种需求就更加突出了。我们建议开发新的计算系统，通过适应其他相关学科（如视觉分析）的最新方法，直接支持数据知情的设计探索。我们将在设计建筑环境和产品的背景下测试和验证我们的解决方案。我的长期目标是开发一种有效地处理替代方案的方法。中期目标是：(1)开发计算方法，通过利用新兴设计数据辅助搜索大量设计方案；(2)识别数据的类型和来源，探索设计师可以使用的可视化数据分析工具；(3)制定工作流程，以便根据数据探索替代方案。我的方法是建立在基于设计的研究基础上的，其中原型系统和实验研究增量和迭代地揭示了新的必要工具特性和用例。在实验阶段，原型将用于产生关于新设计探索系统和数据知情工作流程的可能性的知识。作为一项应用研究计划，该计划旨在为改进的计算设计环境提供更深入的理解和实用的解决方案。好的设计在任何地方都是通往成功的大门，也是创新的关键组成部分。加拿大以服务为中心的经济，可以通过创造新的计算技术，并展示这些技术如何有效地改善结果，在改变设计实践方面发挥强大的领导作用。虽然加拿大的国家竞争力很高，但它在全球市场上的设计竞争力并没有达到我们期望看到的水平。改善这种不平衡的第一步是在创意产业中使用更好的计算方法，包括设计建筑环境。

Designers explore multiple design alternatives. They begin with incomplete, imprecise, and changing design goals, and they distill complex emerging design data into meaningful and pragmatic design ideas for finding 'satisficing' solutions. Here, we use 'design data to refer to any input provided by the external agents to be considered in design-decision making or any output data generated in the process and provided as input back to the design. Current design systems lack support for data-informed design exploration: first, they only support a single-state model, forcing designers to work sequentially. Second, they lack features for dealing with large, diverse, dynamic volumes of design data that grow rapidly as alternatives are explored. These drawbacks are interrelated and must be addressed holistically by improved systems to better support working with alternatives. Thus, this proposal addresses the need for devising new computational design tools and workflows for working with alternatives. This need is more emphasized when large volumes of alternatives become an integral part of design exploration. We propose developing novel computational systems to directly support data-informed design exploration by adapting the state-of-art methods from other related disciplines such as visual analytics. We will test and validate our solutions in the context of designing built environments and products. My long-term goal is to develop a methodology for working effectively with alternatives. The mid-term objectives are: (1) develop computational methods for searching large collections of design alternatives assisted by making use of emerging design data; (2) to identify types and sources of data and explore visual data analysis tools that designers can use; (3) develop workflows for data-informed exploration of alternatives. My approach is built on design-based research where the prototype systems and experimental studies incrementally and iteratively reveal new necessary tool features and use cases. In the experiment phase, the prototypes will be used to generate knowledge about the possibilities for novel design exploration systems and data-informed workflows. As an applied research program, this aims to contribute to a deeper understanding of and practical solutions for improved computational design environments. Good design is everywhere a gateway to success and a key component of innovation. With her service-focused economy, Canada can take a strong lead in changing design practice by creating new computational technologies and demonstrating how these technologies can effectively improve outcomes. Although Canada's national competitiveness is high, its design competitiveness in the global market is not at the levels we expect to see. The first step for improving this imbalance is better computational methods in creative industries, including designing built environments.