Hydraulic Digital Twin
Synthetic, confidentiality-safe workflow for generating hydraulic sensor data, validating signals, detecting anomalies, classifying operating states and producing reports.
Applied AI · research software · engineering data
Applied AI for engineering systems.
I build research-software and machine-learning workflows that connect sensor data, simulation and domain knowledge to monitor, validate and understand complex physical systems.
Selected projects
Four public repositories that show the main shape of my applied-AI and research-software work.
TDMS Sync Checker
Metadata-first QA/QC tool for engineering TDMS files, including timing checks, group/channel synchronisation diagnostics and split-file continuity review.
Structural Audio Anomaly Detection
Audio and signal-processing workflow for structural-test monitoring using time-frequency features, latent representations and anomaly-scoring methods.
LDSFL Meander
Scientific Python implementation of a reduced morphodynamic model for river-centreline evolution, with reproducible examples and GUI/CLI workflows.
Technical focus
- Applied AI: anomaly detection, classification, signal features and model validation.
- Engineering data: sensor synchronisation, experimental workflows, TDMS files and data quality.
- Scientific Python: NumPy, pandas, SciPy, Matplotlib, scikit-learn and reproducible scripts.
- Research software: clear documentation, examples, tests, reports and limitations.
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 data or synthetic data, visual outputs and assumptions.
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 and engineering-data workflows.