Elisa gave her PhD seminar talk titled “Human-AI Interaction and Alignment in Engineering Design: An Exploration of Behavioral, Cognitive, and Neurocognitive Perspectives” (abstract below)!

ABSTRACT: Advances in artificial intelligence (AI) increasingly enable opportunities for providing real-time support to human designers during early-stage design. Novel interactions afforded by these systems (e.g., retrieval of visually similar designs based on sketched inputs) raise fundamental questions regarding their influence on designers’ behaviors, cognitive processes, and design outcomes. For these interactions to be effective, human and AI perspectives should align such that designers’ intentions and expectations are accurately interpreted by AI. My doctoral work explored the interaction and alignment between human and AI systems during design across a series of empirical studies. Through behavioral investigations, I first examined the use of a custom multi-modal AI interface to support concept generation and design space exploration using computational prediction of design behaviors resulting from AI interventions. In a cognitive study on human similarity perceptions, I then compared how humans and AI define similarities between designs, identifying criteria for aligning AI to human representations in design tools. Applying a novel neuroimaging approach, I lastly showed how brain signals during word generation positively align with word representations by large language models (LLMs), indicating the potential for LLMs to directly interface with human thought during design. Across these studies, my work has contributed to understanding and informing effective human interaction with AI in engineering design.

The Co-design lab in front of a projector screen with additional community members appearing on the screen through Zoom. Elisa Kwon is standing at the center.

We are excited to welcome new PhD student Diana Bolaños to the Co-Design Lab. Diana joins the lab from BYU, where she completed her Master’s and Bachelor’s in mechanical engineering. Welcome!

Four members of the Co-design lab in front of a sign reading Welcome to the ASME IDETC-CIE Conference.

Our group is excited to present several papers at the ASME IDETC ‘24 conference! We plan to see everyone in person in Washington DC!

Designing Remote Monitoring for Smart Manufacturing Facilities: Hazard Identification and Classification
Authors: Caseysimone Ballestas, Mansidak Singh, Duy Vu, Kenton Blane Fillingim, and Kosa Goucher-Lambert

Evaluating Design Rationale
Authors: Yakira Mirabito, Xiaowen Liu, and Kosa Goucher-Lambert

Understanding complex sketch recognition strategies for intelligent sketch-based design tools
Authors: Gaëlle Baudoux and Kosa Goucher-Lambert

Human-AI Collaboration Among Engineering and Design Professionals: Three Strategies of Generative AI Use
Authors: Kevin Ma, George Moore, Vikram Shyam, James Villarubia, Kosa Goucher-Lambert, and Eric Reynolds Brubaker

Ananya gave her PhD seminar talk titled “Perceptual Alignment for Design Computing: Quantifying Similarity and Semantic Representations in Early-stage Design” (abstract below)!

ABSTRACT: During early-stage design processes, designers must navigate significant uncertainty and make sense of abstract, multi-dimensional goals (e.g., function, aesthetics, ergonomics), eventually synthesizing them into design outcomes. Data-driven design is a paradigm that aims to leverage data and computational methods to support decision making, allowing designers to surpass cognitive limits (e.g., idea fixation). However, concepts fundamental to decision making during early-stage design (e.g., ‘What are similar design ideas?’ and ‘Will the design reflect dependability?’) are ill-defined, cognitively complex, and not well-represented by computation. Therefore, a key challenge is to align computational representations with how humans perceive and process information, enabling designers to accurately express their intent. To address this challenge, my dissertation research explores behavioral studies and computational techniques to understand and quantify representations (both cognitive and reflected within design artifacts) of these complex, design-relevant concepts. First, I demonstrate how function can be quantifiably compared across engineered systems and products, and how human perceptions of similarity align. Then, I show how intangible semantic prompts (e.g., dependable, versatile, comfortable) can be tangibly reflected in designs, by humans and through human-in-the-loop computation. The insights derived from this work contribute to human-centered computing for early-stage design, enabling designers to more easily and effectively design innovative products.

Yakira gave her PhD seminar talk titled “An Exploration of the Sociotechnical Factors Impacting Design Decisions” (abstract below) and and will be joining MIT as a postdoctoral fellow through their Engineering Excellence postdoctoral program!

ABSTRACT: Errors in design decision making can manifest in many ways; from small misalignments in product affordances, to large scale system failures. In order to improve current design processes, this research aims to better understand the explicit and implicit sociotechnical factors impacting design decisions and associated outcomes. The main research objectives were to (1) investigate the behaviors and cognition influencing decision-making processes, (2) examine the effectiveness of design rationale documentation practices, and (3) analyze the influence of power dynamics on design reviews. To comprehensively assess convergent design decisions, quantitative methods (controlled experiments and computational modeling) and qualitative methods (document analysis, observations, and interviews) were utilized. This mixed-methods approach enabled a thorough investigation of convergent design decisions among students and industry professionals by examining the inputs (information), processes (evaluation and selection), and outcomes (implementation) of these decisions. The results identified decision-making strategies that enhanced design outcomes, developed a framework to represent design rationale more effectively, and articulated the influence of power dynamics in industry design reviews. These contributions provide a transformative approach that empowers engineers to build more innovative and safer products, leveraging design methods and tools informed by a deeper understanding of sociotechnical factors’ influence on design decisions.

Our paper “Semantic properties of word prompts shape design outcomes: understanding the influence of semantic richness and similarity” received a Best Paper Award in Design Cognition at DCC 2024. The paper was led by Ananya Nandy in collaboration with Toyota Research Institute. Way to go! Read the paper here.

Our group is excited to present several papers at Design Computing and Cognition ‘24! We look forward to seeing everyone in Montreal.

Automating Analogical Reasoning: A Wizard of Oz study on the benefits and pitfalls of a sketch-based AI image generator for design
Authors: Gaelle Baudoux and Kosa Goucher-Lambert

Assessing the alignment between word representations in the brain and large language models
Authors: Elisa Kwon, John D. Patterson, Roger E. Beaty, and Kosa Goucher-Lambert

Semantic properties of word prompts shape design outcomes: understanding the influence of semantic richness and similarity Authors: Ananya Nandy, Monica Van, Jonathan Li, Kosa Goucher-Lambert, Matthew Klenk, and Shabnam Hakimi