Flexible optical pressure sensor based on bending-loss-induced intensity modulation with inverse-model-based pressure reconstruction in polymer optical fibers embedded in elastomeric substrates

Abstract

Flexible pressure sensors capable of adapting to non-planar surfaces and enabling spatial pressure mapping are of growing interest across diverse engineering domains. However, conventional electronic sensing approaches often face technical constraints such as electromagnetic interference, wiring complexity, calibration drift, and performance degradation under repeated mechanical loading, particularly when implemented across extended or distributed sensing areas. These challenges have motivated the exploration of alternative transduction mechanisms to improve scalability and signal stability. This thesis presents the design and experimental evaluation of a flexible optical pressure sensing platform based on bending-loss mechanisms in polymer optical fibers, coupled with an inverse-modeling framework for spatial localization and pressure estimation. The developed system consists of flexible elastomeric pads with embedded polymer optical fibers arranged in grid configurations, low-cost optoelectronic acquisition circuits for optical excitation and signal acquisition, and a set of mathematical models that relate measured light intensity variations to applied pressure. The proposed approach integrates an inverse model based on the Moore–Penrose pseudoinverse to determine the location of applied pressure events, along with point-specific estimation models used to reconstruct two-dimensional pressure maps. The system was developed and assessed through a progressive experimental strategy, starting from fundamental sensing configurations and advancing toward a multipoint grid capable of capturing spatial pressure distributions across the sensing area. Experimental characterization was performed under controlled static loading conditions, allowing the quantitative assessment of localization performance, pressure estimation errors, and reconstruction consistency. The findings indicate that the proposed system can generate consistent and interpretable spatial responses under mechanical loading using low-cost materials and optoelectronic components. Localization performance showed stable behavior relative to the grid resolution, while pressure estimation accuracy was strongly dependent on correct spatial localization. Quantitative analyses confirmed the direct relationship between localization error and estimation quality, underscoring the role of spatial mapping in the performance of optical pressure-sensing systems. In this context, the integration of physical modeling with calibrated point-based estimators supported pressure map reconstruction without relying on learning-based approaches. The experimental evaluation also revealed key challenges associated with optical attenuation, electronic noise, and mechanical coupling within the encapsulating substrate, affecting sensitivity and measurement precision, particularly near sensor edges and intermediate locations. Rather than detracting from the overall contribution, these factors help define the operational boundaries of the current prototype and identify critical aspects influencing system behavior. Overall, this work establishes a laboratory-scale proof of concept and an analytical framework for the development and evaluation of flexible optical pressure-sensing systems. The results confirm the feasibility of integrating low-cost optoelectronic acquisition with inverse-model based reconstruction to achieve distributed pressure sensing, while establishing a structured foundation for future improvements in materials, fabrication consistency, and model scalability.

Date	24 Apr 2026
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Bora Ung (Supervisor) & Jean-Marc Lina (Co-supervisor)