Image

Algorithm concept ref [Qin et al. CQG 2025]

Synopsis

This project aims to develop a three-mirror coupled optical cavity system with automated alignment and control. Machine learning will be used to identify optical modes and optimize cavity operation, enabling advanced studies in precision optical control and interferometry.

Description

Optical alignment of the large long baseline suspended mass interferometers is a time consuming task. In the last observation run, 3% time was used in the initial alignment. Fully automated and reliable initial beam steering and optical cavity alignment would be a major stepping stone towards increased duty cycles. This project will develop advanced controls and machine learning (ML) techniques to align complex optical systems like gravitational wave detectors. The project will extend the techniques demonstrated in [Qin et al. CQG 2025] to the complexity of coupled cavities and suspended long baseline cavities.

At ANU, we will build a suspended three-mirror coupled cavity resembling a power-recycled interferometer, using Tip-Tilt suspensions, voice-coil steering mirrors for input beam alignment, and CMOS cameras for beam monitoring. ML algorithms will identify optical modes, automate alignment, and engage length sensing and control to lock the cavity at its operating point. If successful, the ML alignment approach will contribute to enhancing GW detectors’ performance.

Key Learning Outcomes include:

• Hands-on experience with suspended optical cavities and interferometer-like systems.
• Application of machine learning to optical mode recognition and automated alignment.
• Implementation of length sensing and control in complex optical setups.
• Investigation of advanced optical control strategies for high-precision experiments.

Required background

Some knowledge of python or another imperative programming language, image processing experience would be useful. It suits well for Honors and full-time Master project (one semester or two) in 2026.

Research fields