Project Statement
This project focuses not on the style, imagery, or ornamentation of embroidery, but on the essence of embroidery as a structured act.
Embroidery is not merely an accumulation of visual results, but a system composed of a series of irreversible actions, choices, and paths. Every stitch, every turn of the thread, is a trace of decision-making left over time.
This research takes hand embroidery as its core subject, attempting to view it as a structure that can be disassembled, analyzed, and understood, rather than simply a craft or artistic expression. Through continuous recording of the nodes, paths, density, and direction during the embroidery process, this project seeks to return to the act of embroidery itself, understanding "how the work is completed," rather than merely focusing on its finished appearance.
Why Embroidery
The choice of embroidery as the research medium is not based on emotion or cultural symbolism, but on its structural necessity. Embroidery is an irreversible act; once the needle touches the ground, the stitch becomes a physical record of time. Unlike painting or digital images, embroidery does not allow for arbitrary reversals, overwriting, or resetting. Every decision is directly preserved within the work, forming a traceable behavioral trajectory.
Therefore, embroidery provides a remarkably clear research field, allowing behavior, time, and structure to be concretely presented. This project uses embroidery to explore how humans continuously make judgments under limited conditions, and how these judgments are transformed into comprehensible structural information, forming the foundation for subsequent analysis and interdisciplinary understanding.
Core Research Questions
- Can an embroidery work be understood as a structure constituted by behavior, rather than merely a visual image?
- Do the order, position, and direction of needle entry and exit form a describable and analyzable structural language?
- Do the choices and avoidances in the embroidery path reflect the creator's judgments and accumulated experience over time?
- Is there a traceable relationship between stitch density, node distribution, and time consumption?
- When embroidery is transformed into structural data, can AI understand the image result or the human behavioral process?
Method & Documentation
This project uses hand embroidery as the primary behavioral source for research, viewing the embroidery process as an observable and recordable structural generative behavior. The research does not end with the completion of a finished piece, but rather continuously traces back each decision made during the embroidery process, attempting to establish an analyzable recording method at the behavioral level.
To provide a consistent spatial reference for the embroidery behavior, this study employs a fixed-scale Canvas grid system as a benchmark, placing the embroidery actions within a locatable planar coordinate system. Through this system, the positions, directions, and relative distances of the needle's entry and exit points are marked, forming the spatial structural foundation required for subsequent analysis.
The path selection, node formation, and stitch density during the embroidery process are not recorded all at once, but rather gradually established through a multi-layered, iterative process. Each layer of recording deals with only a single structural aspect, avoiding the mixing of behavior, visual elements, and semantics within the same layer, ensuring that the data remains deconstructable for subsequent analysis.
The documentation established in this study does not aim for immediacy or complete reproduction of the workflow. Instead, it focuses on preserving the structural relationships within the embroidery process, ensuring that it remains understandable and reanalyzable even after being separated from the original work. All records are based on human interpretation and labeling, avoiding the simplification of the process's inherent complexity through automation.
Through this approach, this project seeks to establish a sustainable documentation framework between hand embroidery and structural analysis, enabling embroidery to be viewed not merely as an end result, but as a behavioral system that can be repeatedly read and understood.
The Role of AI in This Project
In this project, artificial intelligence is not considered a creator, nor is it intended to replace the act of hand embroidery. AI's role exists solely as an analytical tool for attempting to understand the structure of human behavior after the embroidery process has been transformed into structural data.
This study does not focus on whether AI can generate embroidery images, but rather on whether AI can understand the visual result or the behavioral process itself when embroidery is broken down into structural information such as nodes, paths, directions, and densities. In other words, the function of AI in this project is to test how humans translate highly experiential and embodied manual behavior into a structural language that can be understood by non-humans.
In this process, AI does not access the original work, nor does it participate in the generation of embroidery decisions. All data comes from the results of manual recording and structural organization. AI deals with hierarchically processed behavioral information, not the image or stylistic features of the finished product.
Through structural understanding tests with AI, this project can examine which parts of the embroidery behavior can be formalized, and which still rely on human experience and intuitive judgment. This difference constitutes an important boundary that this research continues to explore.
In this sense, AI is not the core subject of this project, but rather a tool used to reflect the complexity of human behavior, allowing embroidery to be reread and reconsidered in a cross-disciplinary context.
The Role of the Databases
The embroidery structure databases established during this research are an internal working system of the research project and are not part of the content displayed on the website. The databases serve to support structural analysis and cross-layer comparison of embroidery processes, not as publicly accessible or downloadable resources.
These databases are composed of manually recorded and multi-layered organization, covering nodes, paths, directions, densities, and related structural markers in the embroidery process. Their design aims to preserve the structural relationships within the embroidery process, allowing for continued analysis and understanding even after the original work has been removed.
To prevent the structural data from being misused or simplistically interpreted after being removed from the research context, all databases are currently not publicly accessible. This website is only used to illustrate the research direction and methodological framework and does not undertake the function of database display, teaching, or tool services.
The databases themselves will be continuously adjusted and revised as the research progresses. Their content and structure are not prioritized for external consistency but rather for research needs. This choice aims to ensure that the data faithfully reflects the complexity of embroidery processes and is not forced to be simplified due to the need for public access.
In this project, the database is not an outcome, but a continuously operating research device. Its purpose is to support the deepening of structural thinking, not to serve as a display of a completed state.
Closing
This project does not aim for completion or a conclusion, but rather to continuously focus on how the act of embroidery unfolds, is recorded, and understood over time. Through repeated deconstruction and reorganization of the structure, embroidery is temporarily detached from the appearance of the finished product and is re-viewed as a system constituted by actions.
In this process, what is important is not how many results are left behind, but rather preserving a method that can still be read and thought about. What this project attempts to do is to leave understandable structural traces before the manual experience disappears, so that the future may still return to the question of "how it was completed."
The research is ongoing, and the understanding will continue to grow.