MakeAloud: Think-Aloud to Bridge Design-Fabrication Workflows

Ritik Batra, Kendra Wannamaker, George Fitzmaurice, Justin Matejka

January 2026 · Proceedings of the 2026 Designing Interactive Systems Conference (DIS)

Abstract

Translating Computer-Aided Design (CAD) models into physical objects requires expertise and adjustments to navigate fabrication constraints. Makers develop this tacit knowledge by understanding materials, techniques, and practical requirements. Adjustments are typically shared with designer collaborators through sketches and text. However, this documentation lacks situated knowledge gained during fabrication and remains disconnected from the model. To explore how computational tools could address these limitations, we developed MakeAloud, a design probe leveraging AI to capture makers’ in-situ knowledge with hand-tracking hardware and think-aloud computing and then generate design insights within collaborators’ CAD tools. Through a study with woodworkers and designers, we identify three design considerations for designer-maker collaboration tools: surfacing fabrication constraints in CAD to preserve designer intent, supporting asymmetrical domain expertise through AI-mediated communication, and building collective fabrication knowledge archives. This work contributes empirical insights into how AI can bridge design and fabrication workflows, offering pathways for cross-disciplinary collaboration.

DOI PDF

Figures

Figure 1: Workflow illustrating how designers and makers can collaborate using MakeAloud (orange). (a) During the fabrication workflow, the maker (blue) captures their situated knowledge with MakeAloud video recording them and their think-aloud utterances. (b) The designer (green) can then use MakeAloud in Autodesk Fusion to generate insights from the captured fabrication workflow and receive guidance on how to adjust the CAD model based on the maker’s fabrication constraints.

Figure 2: Screenshot of the mobile iOS application for cap- turing fabrication workflows.

Figure 3: Walkthrough of makers and designers collaborating using MakeAloud. (a) First, the maker records themselves during their workflow using our capture system. (b) MakeAloud makes occasional AI-powered interruptions to enrich the captured think-aloud data. (c) The maker can gesture to make forced insights that surface critical considerations to the designer. (d) Once the maker has completed their recording, the designer can view insights generated from the workflow within Fusion. (e) Each insight is attached to the related component of the CAD model. (f) The designer can ask follow-up questions about the maker’s thought process and decision-making during their workflow.

Figure 4: Screenshot of MakeAloud’s Autodesk Fusion add-in, showcasing several insights from a fabrication workflow.

Figure 5: MakeAloud leverages think-aloud data from maker workflows to generate insights in Autodesk Fusion.

Figure 6: Walkthrough demonstrating how RAG (Retrieval Augmented Generation) works in MakeAloud.

Figure 7: Our study matched each designer participant to the captured data of a random maker participant.

Figure 8: We provide the study participants with this flawed 3D model and printed out drawing.

Figure 9: Timeline of maker participants using MakeAloud to document their workflows.

BibTeX

@inproceedings{10.1145/3800645.3813028,
author = {Batra, Ritik and Wannamaker, Kendra and Fitzmaurice, George and Matejka, Justin},
title = {MakeAloud: Think-Aloud to Bridge Design-Fabrication Workflows},
year = {2026},
isbn = {9798400725630},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3800645.3813028},
doi = {10.1145/3800645.3813028},
abstract = {Translating Computer-Aided Design (CAD) models into physical objects requires expertise and adjustments to navigate fabrication constraints. Makers develop this tacit knowledge by understanding materials, techniques, and practical requirements. Adjustments are typically shared with designer collaborators through sketches and text. However, this documentation lacks situated knowledge gained during fabrication and remains disconnected from the model. To explore how computational tools could address these limitations, we developed MakeAloud, a design probe leveraging AI to capture makers’ in-situ knowledge with hand-tracking hardware and think-aloud computing and then generate design insights within collaborators’ CAD tools. Through a study with woodworkers and designers, we identify three design considerations for designer-maker collaboration tools: surfacing fabrication constraints in CAD to preserve designer intent, supporting asymmetrical domain expertise through AI-mediated communication, and building collective fabrication knowledge archives. This work contributes empirical insights into how AI can bridge design and fabrication workflows, offering pathways for cross-disciplinary collaboration.},
booktitle = {Proceedings of the 2026 Designing Interactive Systems Conference},
pages = {2037–2053},
numpages = {17},
keywords = {Design, Fabrication, Documentation, Collaboration, AI},
location = {
},
series = {DIS '26}
}