Force-Position Photo-Robotic Approach for the High-Accurate Micro-Assembly of Photonic Devices

TLDR

Micro‑assembly of photonic elements demands highly accurate free‑space position control and constrained‑space contact detection, making precise measurement a critical challenge. The authors employ Fabry‑Perot interference for 1‑D displacement sensing, integrate it with multi‑DoF robotic motion and a compliant photo‑robotic structure, and introduce a correlation‑based contact detection algorithm to estimate lamella position and control UV‑adhesive thickness during assembly. Using this approach, they achieved 27.6 nm position and 2.1 m° orientation accuracy and detected contact within 80 nm error in assembling an optical lamella onto a fiber.

Abstract

Recently, micro-assembly of individual photonic elements has been highly attracting to provide new optical functionalities and products with high performances. This approach requires position control in free space and contact detection in constrained space. In both cases, getting accurate measurements remains a key lock. For this sake, the FP (Fabry-Perot) interferences happening during the active alignment of optical elements is exploited. This 1D displacement measurement is combined with multi-DoF (Degrees-of-Freedom) robotic motions and a compliant structure along a photo-robotic approach to provide high accurate estimation of both multi-DoF position and contact forces during the assembly process. This approach has been applied to the assembly of an optical lamella with respect to a fiber. An original contact detection algorithm has been proposed, which relies on correlations between the interference figure and the expected irradiance deduced from the nano-positioner internal sensors. This contact detection enables to estimate the position of the lamella with respect to the fiber along the optical axis and thus control the UV-adhesive thickness required to achieve high optical performances after assembly. Experimental investigations using this new approach shows that a positioning accuracy of 27.6 nm and 2.1 m° (standard deviation values) are reached for positions and orientations respectively. The contact detection algorithm has also been studied experimentally and results shows that the contact position can be detected with a maximum error of 80 nm, which