Open Source & Software

Software is a major part of my engineering practice.

I use Python both as a scientific-computing environment and as an engineering automation language for modelling, validation, requirements analysis, data processing, testing, and technical documentation.

I favour open-source software, standard formats, reproducible workflows, and tools that can be inspected and maintained independently of proprietary platforms.

My regular scientific-computing stack includes:

• Python;
• NumPy;
• SciPy;
• pandas;
• scikit-learn;
• statsmodels;
• xarray;
• Matplotlib;
• SymPy;
• PyTorch.

Typical applications include numerical modelling, data analysis, signal processing, state estimation, engineering automation, and model validation.

I use software-engineering practices that are common in production Python projects:

• Git-based version control;
• automated testing with pytest;
• static analysis and linting;
• type checking;
• continuous integration;
• packaging and dependency management;
• reproducible environments;
• code review;
• automated quality checks.

My current toolset includes Git, pytest, Ruff, mypy, uv, Bash, GNU/Linux, Emacs, Org mode, and LaTeX.

My engineering work has also involved:

• IBM DOORS;
• MATLAB and Simulink;
• LabVIEW;
• TestStand;
• ControlBuild;
• LTspice.

Communication interfaces and protocols encountered in validation and test activities include Ethernet, CAN, CANopen, and ARINC 429.

A recurring part of my work is replacing repetitive or fragile manual processing with reviewable software.

Examples include:

• comparison of regulatory requirements with project requirements stored in IBM DOORS;
• Python tooling for railway validation;
• development of the PyMamut test library;
• automated processing and analysis of test data;
• numerical-model testing and regression checks;
• generation of structured technical outputs from machine-readable data.

The objective is not automation for its own sake, but reduction of avoidable manual work while retaining traceability and engineering judgement.

I maintain and contribute to software projects related to scientific computing, engineering, data access, Linux tooling, and terminal applications.

Selected areas include:

• scientific Python;
• PEM fuel-cell modelling;
• public-data tooling;
• Linux and Emacs configuration;
• terminal user interfaces;
• test and CI improvements in existing projects.

My public repositories are available on GitHub.

My preferred environment is based on open tools:

• Debian GNU/Linux;
• Emacs;
• Org mode;
• Git;
• Python;
• Bash;
• standard Unix utilities.

This environment is used both for software development and for technical writing, documentation, publishing, and reproducible research workflows.

I prefer software that is:

• understandable;
• testable;
• documented;
• version-controlled;
• reproducible;
• based on open standards where practical;
• maintainable after the original development context has disappeared.

The same principles apply to engineering scripts and scientific models as to general-purpose software.