11-19 更新 实时
首页/ 深度/ 正文

新型 AI 副驾驶助工程师设计更优结构

深度 来源:MIT News AI 2026-11-13 13:00 0 阅读 机器翻译
新型 AI 副驾驶助工程师设计更优结构

利用算法来设计结构,已经催生出为超轻型飞机、高带宽天线和 F1 赛车优化的新部件。

但最优秀的算法设计技术仍然需要大量计算和时间:用户只有在设计完成后才能看到最终设计,而此后他们往往必须从头重新启动设计流程才能做出修改。这一过程也让人更难确信最终设计在结构上是可靠的。

MIT 的研究人员正在让算法设计变得更具协作性,他们开发了一套像 AI 副驾驶一样运作的系统。在算法生成结构的过程中,它会把各个区域呈现给人类设计师以征求意见。

在一篇描述该系统的开放获取《Communications Engineering》论文中,研究人员表明,与纯算法方法相比,该方法能生成更易制造、屈曲载荷更高的结构,而总设计时间仅增加 15 秒。

研究人员认为,这种方法采用了一种称为拓扑优化的设计技术,有望让更广泛的设计工程师群体更容易使用算法设计,并使其更具可制造性。

“目前,在设计过程推进的过程中,人类没有任何自主权或控制权,”论文第一作者、’26 届博士 Dat Ha 解释道。他不久前刚从 MIT 土木与环境工程系博士毕业,现为 LeMessurier 的结构设计师。“对于这些极其复杂的设计,我们也希望借助我们的大脑——世上最复杂的机器——来以对工业界更实用的方式解决这些设计问题。”

与 Ha 共同参与这篇论文的还有 Ferdous Alam,他曾是 MIT 博士后,现为佐治亚理工学院助理教授;MIT 的 Jerry McAfee (1940) 工程学教授 Markus Buehler;MIT 机械工程副教授 Faez Ahmed;以及资深作者、MIT 土木与环境工程副教授 Josephine Carstensen。

将人类纳入回路

拓扑优化是最先进、最有前景的算法设计技术之一。它使用形式化优化技术,迭代地在给定设计空间内进行填充,以最大程度实现既定目标。对大多数应用而言,这一过程需要大量计算和时间。当最终完成时,设计可能难以制造,或存在工程师未曾考虑的其他问题。例如,一些设计可能在单个点上具有八个或更多梁连接,这在现实世界中很难制造。

“工程师总是非常怀疑计算机程序,特别是当它们给出看起来非常奇怪或出乎意料的设计时,”Ha 说。“从理论上讲,拓扑优化确实非常强大,可以实现出色、高性能的设计,但其黑箱性质使其在工业界未被广泛采用。”

拓扑优化已在汽车和航空航天行业获得关注,但在设计建筑或桥梁方面的应用较少,部分原因是设计工程师需要确信他们可以信任输出结果。

“一个很大的限制因素是,设计工程师对设计如何演进几乎没有影响力,因此无法控制不理想的特性,”Carstensen 解释道。“缺乏控制权和自主权也阻碍了它的使用。”

为了解决这些问题,研究人员让拓扑优化变得更具协作性,Carstensen 将这种方法比作拥有一个设计副驾驶。

此前将人类判断引入其中的尝试会大幅拖慢进度。研究人员的新系统利用机器学习,快速识别出算法设计中人类最可能想要修改的区域。为了训练这个机器学习模型,研究人员构建了一个合成数据集,其中包含针对数千个设计的可能的人类反馈,这些反馈对美学、性能标准和可制造性等方面提出了批评意见。

“我们当时想,如果我们能记录下人类在哪些地方对这些程序做出了修改,那么这些选择或许存在某种规律,而这与其他类型的机器学习方法并没有太大区别,”Ha 说。

Carstensen 认为,这套系统将算法设计与人类设计两种方法各自的优势结合了起来。

“其思路是挖掘设计师的大脑,让他们随着设计的演进注入自己的领域知识,”Carstensen 说。

在演示中,研究人员展示了他们的“人在回路”方法将可制造性和线性屈曲载荷提升了 39%,而总设计时间仅略有增加。

目前,该方法只允许用户在已识别出的区域内把结构做得更厚或更薄,但他们希望在未来版本中加入其他修改选项。

“我们希望,通过解决这些障碍,能让拓扑优化面向更广泛的受众,不仅是工业界,也包括学术界的研究人员和学生,”Carstensen 说。

从设计到结构

研究人员认为,他们的方法可以补充纯 AI 的设计方法。

“我们认为两者都有存在空间,”Carstensen 说。“如果你所在的学科拥有大量计算资源,你可以使用另一种方法。但如果你确实需要人的输入,就可以采用这种协作式方法。”

加入人的输入也有助于提升信任。

“在某些场景下,人类需要在推进最终设计之前做出最终决定,”Carstensen 说。“通过让人类参与进来,我们可以把性能更高的设计推向现实世界,从而创造出性能更高的结构。因为这不仅仅是设计出好的制品,而是创造出人类真正会接受并制造出来的好制品。”

这项工作部分得到了 MathWorks 工程 fellowship 基金、TATA Steel 以及 MIT 生成式 AI 影响力联盟(MIT Generative AI Impact Consortium)的支持。

译文由英文原文机器翻译,术语按行业惯例统一,可在文末展开原文核对。

查看英文原文(New AI “copilot” helps engineers design better structures)

The use of algorithms to design structures has led to new parts optimized for ultralight aircraft, high-bandwidth antennas, and F1 racing cars.

But the best algorithmic design techniques still require lots of computation and time: Users only see the final design after it’s complete, after which they often must start the design process all over again to make changes. That process also makes it harder to trust that the final designs are structurally sound.

MIT researchers are making algorithmic designs more collaborative with a system that acts like an AI copilot. As the algorithm is generating the structure, it presents regions to the human designer for feedback.

In an open-access Communications Engineering paper describing the system, the researchers showed it led to more easily manufacturable structures with higher buckling loads while adding only 15 seconds to the total design time compared to all-algorithmic approaches.

The researchers believe the approach, which uses a design technique known as topology optimization, could make algorithmic design more accessible and manufacturable for a broader class of design engineers.

“Right now, the human doesn’t have any agency or control over the design process as it’s playing out,” explains lead author Dat Ha PhD ’26, a recent MIT civil and environmental engineering doctoral student who is now a structural designer at LeMessurier. “For these really complex designs, we wanted to also tap into our brain, which is the most complex machine there is, to solve these design problems in a way that’s more practical to industry.”

Joining Ha on the paper are Ferdous Alam, a former MIT postdoc who is now an assistant professor at Georgia Tech; Markus Buehler, the MIT Jerry McAfee (1940) Professor of Engineering; Faez Ahmed, MIT associate professor of mechanical engineering; and senior author Josephine Carstensen, MIT associate professor of civil and environmental engineering.

Bringing a human in the loop

Topology optimization is one of the most advanced and promising algorithmic design techniques. It uses formal optimization techniques to iteratively fill in the space given for a design in a way that maximally achieves the stated goals. For most applications, the process requires a lot of computation and time. When it’s finally done, the designs can be impractical to build or have other issues the engineer didn’t consider. For instance, some designs might have eight or more beam connections at a single point, which would be difficult to make in the real world.

“Engineers are always very skeptical of computer programs, especially if they give you very weird-looking or unexpected designs,” Ha says. “On paper, topology optimization is really powerful and can achieve great, high-performing designs, but its black box nature has made it less widely adopted in industry.”

Topology optimization has gotten traction in the automotive and aerospace industries, but it’s been used less to design buildings or bridges, partly because design engineers need to know they can trust the output.

“A big limiting factor has been that the design engineer has very little influence over how the design evolves and therefore can’t control undesirable features,” Carstensen explains. “The lack of control and lack of ownership also inhibits its use.”

To address those problems, the researchers made topology optimization more collaborative in an approach that Carstensen compares to having a design copilot.

Prior attempts at adding human judgment slowed things down dramatically. The researchers’ new system uses machine learning to quickly identify the most likely regions of an algorithmic design that a human would want to change. To train the machine-learning model, the researchers created a synthetic dataset of probable human feedback to thousands of designs that critiqued things like aesthetics, performance criteria, and manufacturability.

“We thought if we could create a record of where humans have made changes to these programs, then perhaps there’s a pattern to those choices, which is not very different from other types of machine-learning approaches,” Ha says.

Carstensen sees the system as bringing together the best of algorithmic and human design approaches.

“The idea is to tap into the brains of the designers and get them to infuse their domain knowledge as the design evolves,” Carstensen says.

In demonstrations, the researchers showed their human-in-the-loop approach improved manufacturability and linear buckling load by 39 percent while only modestly increasing the total design time.

Right now, the approach only allows users to make structures within identified regions thicker or thinner, but they hope to add other options for modification in future versions.

“The hope is that by addressing these hurdles, we open up topology optimization to a much wider audience, not only in industry but also to researchers and students in academia,” Carstensen says.

From designs to structures

The researchers see their approach complementing AI only-approaches to design.

“We are arguing there’s room for both,” Carstensen says. “If you are in a discipline that has a lot of computational resources, you can use the other approach. But if you really need the human input, you can use this collaborative approach.”

Adding human input also serves to increase trust.

“In some settings, the humans need to make the final decisions before advancing the final design,” Carstensen says. “By involving humans, we can get higher-performing designs to the real world to create higher-performing structures. Because it’s not just about designing good artifacts; it’s about creating good artifacts that humans will actually accept and make.”

The work was supported, in part, by the MathWorks Engineering Fellowship Fund, TATA Steel, and the MIT Generative AI Impact Consortium.

正文保留英文原文,标题与摘要由机器翻译。

本文由采集器自动同步

原文发布于 MIT News AI(New AI “copilot” helps engineers design better structures),标题、摘要与正文由机器翻译,术语按行业惯例统一,未经人工逐句校订。

阅读原文 ↗