Tidal Blade Test Analysis
Public-safe structural-test workflows for full-scale tidal blade data: TDMS inspection, static response, fatigue-cycle summaries, natural-frequency helpers and applied-AI screening.
Applied AI · research software · engineering data
Applied AI for sensor-rich engineering and environmental systems.
I build public-safe research-software and machine-learning workflows that connect sensor data, simulation and domain knowledge for monitoring, validation, anomaly detection and engineering decision support.
Interactive demos
Two projects can now be explored directly in the browser without cloning a repository or setting up a local Conda environment.
The demos use public-safe synthetic or example data, so they show the workflow while protecting private engineering and monitoring records.
Portfolio focus
These projects are selected to show applied AI beyond model fitting: data quality, reproducibility, diagnostics, reporting, domain constraints and safe publication boundaries.
Selected projects
Public repositories showing applied-AI, research-software, structural-testing and environmental-monitoring workflows.
Urban Drainage Sensor Data Toolkit
Public-safe urban drainage telemetry QA/QC with synthetic monitoring data, static reports, anomaly screening and synthetic monitoring-map outputs.
Meander Morphology Classifier
Scientific-ML toolkit and Streamlit GUI for curvature-based meander classification using CWT spectra, autoencoder latent spaces and clustering.
TDMS Sync Checker
Engineering-data QA/QC tool for TDMS files, with timing metadata inspection, group/channel synchronisation review and continuity diagnostics.
Hydraulic Digital Twin
Confidentiality-safe digital-twin demonstration using synthetic sensor data, validation checks, anomaly detection, operating-state classification and automated reports.
Structural Audio Anomaly Detection
Signal-processing and ML workflow for structural-test monitoring, including feature extraction, similarity scores and anomaly labels.
Technical focus
- Structural test data: TDMS inspection, natural-frequency helpers, static response and fatigue-cycle summaries.
- Engineering and environmental telemetry: timestamp cleaning, missing-data checks, daily summaries and reports.
- Applied AI: anomaly detection, robust screening, signal features and model validation.
- Scientific ML: autoencoders, latent spaces, clustering and spectral features.
- Research software: documented examples, tests, synthetic data and reproducible outputs.
Repository style
I try to make repositories useful as engineering artefacts, not only as code. Where possible, projects include a clear problem statement, quick-start instructions, example or synthetic data, visual outputs, assumptions and limitations.
This portfolio emphasises reproducible workflows and honest boundaries around data, models and confidentiality.
Contact
I am interested in applied AI, research software, digital twins, anomaly detection, engineering-data workflows and environmental monitoring.