Interacting with Design Alternatives

该提案提出了一个问题：计算设计系统如何更好地支持设计工作？具体来说，它提出设计工作可以通过显式表示和与替代解决方案的交互而显著地实现。为什么这很重要？因为设计师通过创造和考虑替代方案来工作。然而，现有的计算设计工具几乎没有提供这样的功能，因此不得不采用既无效又低效的笨拙解决方案。由于两个原因，拟议的研究将使用一种称为参数化建模系统（PCAD）的计算设计工具。首先，这些工具处于设计工具开发的前沿，因此世界上有熟练的用户使用它们。其次，这些工具提供了清晰、简单和有效的方式来表示备选方案。虽然现有的PCAD系统理论上定义了替代方案，但它们的接口通常提供对设计的串行访问。我的目标是改变PCAD接口的这一特性，以支持使用替代方案进行探索。我之前的工作已经发现了与设计备选方案一起工作的表示和交互技术，并定义了本提案的目标。该领域的大多数研究都采用自动机制来生成替代方案，尽管众所周知的用户偏好是主动和直接参与选择。这里提出的研究认为，用户与可能存在的大空间的直接交互提出了一个必须解决的困难和具有挑战性的研究和设计问题，即使在自动设计生成的背景下也是如此。因此，我们的研究计划是设计、实施和验证几个原型系统，以便与设计替代方案一起工作。有了每一个，我们就能更多地了解在支持设计替代方案时什么可行（什么不可行）。研究计划包括三个步骤：(1)设计交互原型的设计方案，咨询专业用户。(2)实现，将交互原型与当代参数化工具相结合。(3)在现实世界（设计办公室）和模拟现实世界环境（如专业办公室和扩展工作坊）中进行评估。这项研究的最终目的是通过实现一次尝试多个想法的通用设计策略，改善计算机支持设计师的方式。

This proposal asks the question: How can computational design systems better support design work? Specifically, it proposes that design work can be significantly enabled by explicit representation of and interaction with alternative solutions. Why is this important? Because designers work by creating and considering alternatives. Yet available computational design tools provide almost no such functionality, thus forcing awkward workarounds that are both ineffective and inefficient.The proposed research will work with a kind of computational design tool known as parametric modeling systems (PCAD), for two reasons. First, such tools are at the forefront of design tool development, so there are skilled users in the world working with them. Second, such tools provide clear, simple and effective ways to represent alternatives. While extant PCAD systems theoretically define alternatives, their interfaces generally provide access to designs serially. My goal is to change this feature of PCAD interfaces to support exploration using alternatives.My prior work has discovered representations and interactive techniques for working with design alternatives, and has defined the objectives for this proposal. Most research in the area employs automatic mechanisms for generating alternatives, in spite of well-known user preferences for actively and directly engaging with choices. The research proposed here takes the view that direct user interaction with possibly large spaces of alternatives presents a difficult and challenging research and design problem that must be addressed, even in the context of automatic design generation. Our research program is thus to design, implement and validate several prototype systems for working with design alternatives. With each, we understand more about what works (and what does not work) in supporting design alternatives.The research program comprises three steps: (1) Design interactive prototypes for design alternatives, consulting with professional users. (2) Implement, combining interactive prototypes with contemporary parametric tools. (3) Evaluate in real-world (design offices) and simulated real-world contexts (such as professional offices and extended workshops).The proposed research ultimately aims to improve the ways that computers support designers by enabling the universal design strategy of trying more than one idea at a time.