University of Victoria
Research Assistant · Nanophotonics Research
This work combined computational nanophotonics research, COMSOL-based electromagnetic simulation, scientific data processing, research workflow automation, technical knowledge management, and semantic retrieval with AI-assisted research search. I designed and completed an Engineering Simulation Workflow Platform to centralize simulation data, technical documentation, and research records.
- Role
- Research Assistant
- Dates
- Dates will be added.
- Location
- Victoria, BC
- Python
- MATLAB
- COMSOL Multiphysics
- React
- Docker
- AWS
- GitHub Actions
- YAML
01
Background
Research focus
The research modelled how surface roughness on whispering-gallery-mode microdisk cavities scatters light and degrades optical performance — work that depended on running and interpreting large numbers of COMSOL electromagnetic simulations.
Original workflow
Simulation inputs, output files, plots, technical notes, and research records were distributed across different tools and directories. COMSOL simulations generated complex scientific datasets, and reproducing a previous experiment meant manually reconnecting simulation parameters, files, analysis scripts, and documentation. Locating a past configuration or result depended heavily on manual data organization.
02
Research Problem
- Scientific Research Challenge
- Understand how sidewall and top-surface roughness affects scattering loss and the quality factor of whispering-gallery-mode optical microcavities.
- Engineering Workflow Challenge
- Create a structured and reproducible system for managing simulation configurations, generated datasets, plots, technical documentation, and historical research results.
03
Engineering Simulation Workflow Platform
I designed and completed a reusable engineering platform that centralized the research group's simulation configurations, COMSOL-generated results, processed scientific datasets, plots and exported files, experiment metadata, technical documentation, and historical research records.
- Simulation configurations and COMSOL-generated results
- Processed scientific datasets, plots, and exported files
- Experiment metadata and technical documentation
- Historical research records
- Automated data ingestion
- Semantic retrieval
- AI-assisted search
Status
The platform has been completed. The codebase is currently being refined and documented before public release.
04
My Role
I was responsible for the complete design and implementation of the platform, connecting its architecture directly to the requirements discovered during the optical microcavity research.
- Analyzed the original research and simulation workflow.
- Identified data-management and reproducibility problems.
- Designed the full system architecture.
- Defined the research data model.
- Designed and implemented automated ingestion pipelines.
- Built data-processing workflows for simulation outputs.
- Implemented semantic retrieval and AI-assisted search workflows.
- Integrated simulation data, documentation, and research records into one platform.
- Containerized the application and its supporting services.
- Established reproducible deployment and configuration practices.
- Maintained data processing and application infrastructure using Docker, AWS, GitHub Actions, and YAML.
05
Platform Architecture
Scientific simulation outputs feed directly into the platform's ingestion and retrieval pipeline.
Infrastructure & deployment
- Docker
- AWS
- GitHub Actions
- YAML
06
Simulation and Research Workflow
Each run generated Gaussian-distributed surface profiles using a controlled root-mean-square roughness height and correlation length, so results could be compared consistently across configurations.
- Stage 01Define Geometry and Optical Parameters
- Stage 02Generate Statistical Surface Roughness
- Stage 03Run 3D COMSOL Simulation
- Stage 04Extract Electromagnetic and Scattering Results
- Stage 05Process and Compare Data
- Stage 06Store Results, Documentation, and Research Records
- Stage 07Retrieve Historical Knowledge Through Search
07
Platform Outcome
- Centralized simulation data, documentation, and research records.
- Reduced reliance on manual file organization.
- Improved reproducibility of engineering workflows.
- Enabled structured access to historical simulation results.
- Supported automated data ingestion and processing.
- Enabled semantic retrieval and AI-assisted research search.
- Established a reusable engineering workflow for future simulation projects.
- Reduced manual data-management effort by approximately 40% by centralizing simulation data, technical documentation, and research records.
08
Research Outcome
The completed research produced a three-dimensional full-wave electromagnetic modelling framework for studying surface-roughness-induced scattering in whispering-gallery-mode microdisk cavities.
- Surface roughness degrades optical cavity quality factor.
- The research examined roughness on multiple cavity surfaces.
- Statistical roughness profiles were generated with controlled parameters.
- The modelling approach enabled detailed analysis of scattering effects.
- The results were compared with established semi-analytical models.
- The work provides guidance for fabrication optimization and future predictive modelling.
09
Published Research
University of Victoria
Optical Microcavity Surface Roughness Modelling
Heming Qin
A three-dimensional full-wave electromagnetic modelling framework implemented in COMSOL Multiphysics to investigate how surface roughness affects scattering loss and the quality factor of whispering-gallery-mode microdisk cavities.