Development of an Optical Force Sensor: A Novel Approach for Monitoring Physical Interaction in Robotic Walkers
Name: DANIEL EDUARDO GARCIA ALVAREZ
Publication date: 20/02/2026
Examining board:
Name |
Role |
|---|---|
| ANSELMO FRIZERA NETO | Coorientador |
| CAMILO ARTURO RODRIGUEZ DIAZ | Presidente |
| EDUARDO ROCON DE LIMA | Examinador Externo |
| MARCELA CRISTINA MUNERA RAMÍREZ | Coorientador |
| MARIANA LYRA SILVEIRA | Examinador Interno |
Summary: In recent years, demographic changes, population aging, and the increasing prevalence of neurological and musculoskeletal disorders have contributed to a growing number of individuals with mobility impairments. As a result, gait assistive technologies have become an important research focus, with smart walkers (SWs) emerging as a promising solution to provide physical, cognitive, and sensory support during rehabilitation and daily living activities. A key requirement for these devices is the accurate estimation of user–walker interaction forces, which constitute the primary communication channel for intention detection, adaptive assistance, and gait analysis. Conventional force-sensing technologies commonly used in SWs, such as load cells, force-sensing resistors, and high-resolution multi-axis sensors, present several limitations, including mechanical fragility, sensitivity to misalignment, complex integration, and high production costs. These drawbacks restrict their scalability and long-term deployment in real-world scenarios. To address these challenges, this work proposes the design, development, and validation of an optical force sensor based on a waveguide sensing principle. The sensor integrates light emitters and photodiodes embedded within a silicone layer, where contact-induced surface deformations modulate internal light propagation. The proposed sensing system was experimentally characterized through controlled indentation tests and validated using a reference load cell. In addition, a feedforward neural network was employed to model the nonlinear relationship between the optical signals and the applied force. Furthermore, multiple sensor configurations were evaluated to analyze the influence of encapsulation material and illumination wavelength on sensor performance. The proposed solution was also directly compared with a commercial distributed force-sensing system under equivalent loading conditions. Finally, the optical sensor was integrated into a socially assistive walker and assessed during path-following tasks with healthy participants. Experimental results demonstrate that the proposed sensing approach can accurately estimate user–walker interaction forces. Overall, this dissertation demonstrates the feasibility of waveguide-based optical sensors combined with data-driven models as a robust and scalable approach for force estimation in SWs, enabling reliable monitoring of user–device physical interaction.
