MATLAB & Simulink
Control systems, power systems, power electronics, signal processing and custom solvers.
- Model building and debugging
- Convergence and stability issues
- Parameter sweeps and case studies
- Result export for publication
Home / Research / Technical
Simulation · Data · Instrumentation
Models that run, models that generalise, and benches that measure.
MATLAB, Simulink, Python, machine learning, optimisation, LabVIEW, data acquisition and reproducible research workflows.
Simulation & modelling
Built to be defended, not just to produce a plot.
Python engineering
NumPy, SciPy, Pandas and simulation loops written so somebody else can run them next year.
- Clean, commented, runnable scripts
- Reproducible environments
- Automated figure and table export
- Version-safe file structure
Optimisation
Objective formulation, constraint handling and solver selection — with a sanity check on the result.
- Single and multi-objective problems
- Metaheuristics and classical solvers
- Benchmark validation
- Pareto front analysis
Machine learning & data
The part reviewers attack first.
Most ML papers are rejected on evaluation, not on the model. We spend the effort where it is scrutinised.
Model development
Baselines first, then the complicated thing — so the gain is measurable and attributable.
Honest evaluation
Chronological and grouped splits, no leakage from resampling, and metrics that match the decision being made.
Uncertainty & calibration
Conformal prediction, prediction intervals and calibration curves, explained in plain terms.
Reproducibility
A notebook or script that regenerates every number and figure in the manuscript from the raw data.
A warning worth paying for: if the results look too good, we will say so and go find the leak before a reviewer does. That conversation is uncomfortable once and useful forever.
LabVIEW & instrumentation
Measurement systems and the records that make them credible.
VI development
State machines, producer-consumer architectures, front-panel design and error handling.
DAQ & hardware
NI USB, myRIO and cRIO configuration, signal conditioning, sampling and sensor interfacing.
Test benches
Hardware-in-the-loop setups, automated test sequences, logging and calibration documentation.
Also on this bench
ETAPPSCADPSIMPLECSProteusMultisimOpenDSSOriginMinitabPyomoGAMSCVX
Next step
Send the model, the data or the error message.
A stuck simulation, an unbelievable accuracy score or an empty measurement log — all three are normal starting points.