Planar Position Measurement of a Flat Object Based on Distributed Fiber Optic Sensors

Abstract

In the semiconductor and photovoltaic industries, contactless positioning systems are increasingly employed to handle ultra-thin wafers, minimizing the risk of mechanical damage. A suitable and integrated sensor system would make this handling method more attractive in future applications. This Msc. thesis explores a novel concept in planar position measurement for flat, reflective objects using fiber optic sensors distributed across a measurement plane. This sensor configuration creates light profiles via transmitting fibers and utilizes the reflected light detected by receiving fibers to localize the object's position.
A Gaussian-distributed intensity pattern is first applied to the light profiles to validate this concept, establishing the theoretical relationship between brightness and planar positions. A prototype with a single set of fibers is then constructed to verify the proposed brightness model and determine the optimal, sparsest fiber spacing that maintains adequate resolution. In the next phase, algorithms are developed using simulated sensor arrays to calculate planar degrees of freedom based on the actual positions and brightness values from receiving fibers near the object's boundary. These algorithms are then implemented in a fabricated sensor array prototype, forming a complete distributed fiber optic sensor system.
The sensor's performance is evaluated using samples with varying center positions and orientations, resulting in an average error of less than 1 mm in the in-plane directions. In conclusion, the proposed sensor configuration limits planar position errors to sub-millimeter accuracy. The measurement range is significantly extended, and the sensor element no longer needs to be attached to the measured object. As a result, this system offers promising potential for contactless planar positioning of thin, fragile products with specular surfaces.