Exploring Human-AI Collaboration in E-Textile Design: A Case Study on Flex Sensor Placement for Shoulder Motion Detection

Abstract

Flex sensors are widely used in e-textiles for detecting joint motions and, subsequently, full-body movements. A critical initial step in utilizing these sensors is determining the optimal placement on the body to accurately capture human motions. This task requires a combination of expertise in fields such as anatomy, biomechanics, and textile design, which is seldom found in a single practitioner. Generative AI, such as Large Language Models (LLMs), has recently shown promise in facilitating design. However, to our knowledge, the extent to which LLMs can aid in the e-textile design process remains largely unexplored in the literature. To address this open question, we conducted a case study focusing on shoulder motion detection using flex sensors. We enlisted three human designers to participate in an experiment involving human-AI collaborative design. We examined design efficiency across three scenarios: designs produced by LLMs alone, by humans alone, and through collaboration between LLMs and human designers. Our quantitative and qualitative analyses revealed an intriguing relationship between expertise and outcomes: the least experienced human designer achieved continuous improvement through collaboration, ultimately matching the best performance achieved by humans alone, whereas the most experienced human designer experienced a decline in performance. Additionally, the effectiveness of human-AI collaboration is affected by the granularity of feedback - incremental adjustments outperformed sweeping redesigns - and the level of abstraction, with observation-oriented feedback producing better outcomes than prescriptive anatomical directives. These findings offer valuable insights into the opportunities and challenges associated with human-AI collaborative e-textile design.

Figures (67)

• Figure 1: Iterative design process of e-textile-based motion capture systems. Each phase is annotated with the domain knowledge it requires. Below the workflow, we map opportunities for AI assistance: LLM-based approaches (blue) can support Phases 1, 4, and 5, while simulation-driven approaches (orange) primarily address Phases 2 and 3. This work focuses on the LLM-assisted stages.
• Figure 2: Overview of the user study protocol. The study proceeds through three phases: Onboarding, Experiment 1 (initial design comparison, addressing RQ1), and Experiment 2 (iterative design refinement comparison across H-Series and LH-Series, addressing RQ2 and RQ3). Within Experiment 2, the H-Series (amber) follows a human-only iteration loop, while the LH-Series (blue) introduces human-AI collaborative design refinements.
• Figure 3: Overview of the technical implementation: LLM-based N-to-1 synthesis (blue) produces $L_0$ during Experiment 1. Within the Feedback & Refinement module, there are two pathways: human designer-only diagnosis and direct design refinement (amber; H-Series), and LLM-mediated two-step refinement informed by structured human feedback (blue; LH-Series). Dashed arrows indicate the iteration loop (up to two cycles). The Rapid Validation pipeline (green) is utilized for evaluating both initial designs and refined ones.
• Figure 4: Data collection setup. (a) Hardware overview: four resistive flex sensors are attached to a compression shirt worn by the subject; a client microcontroller reads the sensor signals and transmits them to a server laptop, while an RGB webcam captures reference video. (b) Close-up of a SpectraFlex™ flex sensor with a custom-fitted fabric container.
• Figure 5: Illustration of the six standardized motions used for performance evaluation.