Soft and stretchable sensors for soft data gloves
| dc.contributor.author | Zhang, Yiqi | |
| dc.date.accessioned | 2026-08-31T17:58:14Z | |
| dc.date.issued | 2026-08-31 | |
| dc.date.submitted | 2026-08-26 | |
| dc.description.abstract | Data gloves are wearable electronic devices equipped with multiple sensors to quantify hand kinematics and contact forces with high fidelity, enabling applications in rehabilitation monitoring, assistive technology, and human-machine interaction. Despite substantial progress, existing off-the-shelf solutions face critical barriers to clinical translation. Rigid or bulky sensing architectures compromise wearability and user comfort, while standardized one-size-fits-all designs lead to sensor misalignment across diverse hand geometries, degrading measurement accuracy. Furthermore, calibration procedures are often time-consuming and impractical for routine or unsupervised use. These limitations are especially pronounced in ambulatory monitoring scenarios, where devices must function reliably across repeated donning and doffing cycles, and accommodate significant inter-user variability in hand size, shape and biomechanics. This thesis addresses the key limitations of existing wearable data gloves through three interrelated technical contributions spanning sensing, modeling, and system-level personalization. First, soft pressure-sensing approaches for palmar contact measurement are developed with a focus on seamless integration into wearable glove platforms. Three sensing modalities are systematically investigated including microwave-resonant sensing, capacitive sensing, and iontronic electric-double-layer sensing. These approaches are comparatively evaluated in terms of sensitivity, mechanical compliance, integration complexity, and suitability for ambulatory use. Based on this analysis, a microstructured soft-stack capacitive architecture is selected for further development. To enable rational design of the microstructure, where performance is governed by nonlinear deformation and evolving contact mechanics, a mechanics-informed finite element modeling framework (FEniCSx/DOLFINx) is established. This is coupled with surrogate modeling to efficiently explore a high-dimensional geometric design space, significantly reducing reliance on empirical, trial-and-error prototyping. Beyond higher sensitivity, this design strategy also enables a more linear response that could help reduce calibration burden by simplifying the mapping between sensor output and applied load. Second, a flexible, thin pillar-structured capacitive bending sensor is proposed for wide-range joint angle measurement. The sensor operates by transducing bending-induced deformation into capacitance change through controlled volumetric redistribution of air gaps within a microstructured dielectric layer, producing an angle-dominant response that is intrinsically less sensitive to placement variability and therefore which requires less calibration than conventional strain-dominant bending sensors, whose outputs vary with fit, pretension, and placement. An analytical capacitance model, supported by finite element parametric studies, informs the optimization of pillar geometry and material properties. The sensing performance is validated through multi-participant gesture recognition experiments using an instrumented data glove with multiple bending sensor integrated, as well as through knee joint angle tracking benchmarked against a motion capture reference. Third, an anatomy-to-layout personalization pipeline is proposed to bridge the gap between sensor-level performance and glove-level usability. The pipeline utilizes user-specific hand measurements, obtained via calibrated imaging or manual landmark identification, to generate parametric glove patterns that align sensing elements with anatomically referenced locations. By anchoring sensor placement to joint axes and contact centroids rather than fixed garment coordinates, the approach reduces misalignment arising from donning variability and inter-user anatomical differences, thereby improving measurement consistency across sessions without imposing extensive recalibration. Collectively, these contributions establish a cohesive design framework for personalized, sensing-integrated data gloves, advancing the state of the art toward practical, reliable, and scalable solutions for on-the-go hand function monitoring. | |
| dc.identifier.uri | https://hdl.handle.net/10012/24138 | |
| dc.language.iso | en | |
| dc.pending | false | |
| dc.publisher | University of Waterloo | en |
| dc.title | Soft and stretchable sensors for soft data gloves | |
| dc.type | Doctoral Thesis | |
| uws-etd.degree | Doctor of Philosophy | |
| uws-etd.degree.department | Mechanical and Mechatronics Engineering | |
| uws-etd.degree.discipline | Mechanical Engineering | |
| uws-etd.degree.grantor | University of Waterloo | en |
| uws-etd.embargo.terms | 0 | |
| uws.contributor.advisor | John , Wen | |
| uws.contributor.advisor | Carolyn , Ren | |
| uws.contributor.affiliation1 | Faculty of Engineering | |
| uws.peerReviewStatus | Unreviewed | en |
| uws.published.city | Waterloo | en |
| uws.published.country | Canada | en |
| uws.published.province | Ontario | en |
| uws.scholarLevel | Graduate | en |
| uws.typeOfResource | Text | en |