11/12/2025
By Opeyemi Akanbi

The Kennedy College of Sciences, Department of Physics, invites you to attend a Master of Science in Physics thesis defense by Opeyemi Samson Akanbi on "Unidirectional Diffraction-Based Strain Sensor for Motion Tracking."

Candidate Name: Opeyemi Samson Akanbi
Defense Date: Friday, November 21, 2025
Time: 8:30 to 10 a.m.
Thesis Title: Unidirectional Diffraction-Based Strain Sensor for Motion Tracking
Location: Olney 136

Committee Members:

  • Wei Guo (Advisor), Kennedy College of Sciences, UMass Lowell
  • Viktor Podolskiy (Member), Kennedy College of Sciences, UMass Lowell
  • Hugo Ribeiro (Member), Kennedy College of Sciences, UMass Lowell

Abstract:
Accurate, stable measurement of large, body-scale deformations remains a core challenge for wearable systems and soft-structure monitoring. Electrical strain sensors (resistive, capacitive, and piezoelectric) can be thin and highly stretchable but often trade sensitivity for hysteresis and suffer from cross-axis coupling under realistic, multi-axial motion. Optical approaches offer geometric, EMI-immune readouts but are rarely engineered for compact, direction-selective wearables. This thesis introduces a unidirectional, diffraction-imaging strain sensor that converts deformation directly into image-space displacements using a transmission grating, a miniature light source, and a chip-scale camera. Building on diffraction-assisted image correlation (DAIC), the optical stack and grating orientation are co-designed so that the correlation metric is maximally responsive to longitudinal strain while rejecting transverse components, thereby providing directional selectivity by geometry rather than material anisotropy. The system integrates a VCSEL emitter and a compact CMOS module within a lightweight package suitable for skin-mounted use. Under tensile loading, deformation of the compliant substrate produces deterministic shifts in the diffracted pattern, which are tracked with sub-pixel precision to recover displacement and strain. A staged validation is presented: (i) benchtop characterization to quantify linearity, resolution, noise, and cyclic stability over small and large strains; (ii) directionality tests applying pure transverse strain to verify suppression; and (iii) on-body studies at the wrist, thigh, and neck to assess performance under natural kinematics. Across tests, the device sustains large working strains (on the order of 10–360%), exhibits low hysteresis and repeatable strain–signal behavior over repeated cycles, and demonstrates robust unidirectionality (minimal response to transverse loading). In human studies, the sensor resolves joint motions with strain envelopes consistent with expected biomechanics—for example, average wrist strains of ~9% during flexion (with higher transient peaks) and ~4% during extension. The contributions are threefold: (1) a compact, wearable-ready diffractive architecture that delivers axis-selective strain sensing; (2) a direct, geometric metrology pipeline enabling sub-pixel displacement resolution without reliance on drift-prone resistive mechanisms; and (3) systematic evidence of low-hysteresis, high-range performance in both benchtop and on-body scenarios. The approach provides a practical foundation for reliable motion capture in human–machine interfaces, rehabilitation monitoring, soft robotics, and structural health monitoring, and suggests clear pathways to multi-modal integration and scalable manufacturing for real-world deployment.