It is now 85 seconds to midnight; giving more reason than ever to learn with urgency, progress with purpose and live deliberately.
I'm a 3rd-year MEng Mechanical Engineering student at the University of Exeter.
I’m interested in simulation-led engineering (CFD/FEA/control), numerical methods, and practical design for manufacture work that actually ships.
I’m also interested in data-centric engineering as a future study focus: designing scalable, reliable data systems for quantitative finance, including streaming market data, low-latency pipelines,
data validation/monitoring, labeling & feature stores, reproducible research (experiment tracking), and deployment workflows that turn raw data into tradable signals.
Some of the repositories here are either:
- coursework/practical submissions exported from Jupyter/MATLAB, or
- cleaned scripts extracted from those notebooks (so the structure can be a bit “cell-by-cell”; not a polished package)
I keep them public as a transparent record of my work and progress.
- Control systems (MATLAB): transfer functions, feedback loops, pole–zero maps, stability checks, step/impulse/ramp responses
- Numerical methods (Python): ODE simulation (Euler, RK2/Heun,
odeint), timestep studies, error analysis, validation - Machine learning (Python): MLPs/CNNs for classification, hyperparameter sweeps, evaluation + visualisations
- Engineering design: manufacturability-focused thinking, iterative build/test, clear reporting
- CFD fundamentals: boundary-layer modelling and similarity solutions (e.g., Blasius)
-
Control-Systems-Analysis-in-MATLAB
Closed-loop analysis, stability assessment, pole–zero maps, and time-domain responses for multiple feedback systems. -
Mass-Spring-Damper-Simulation-Euler-vs-Heun-vs-odeint
Coupled mechanical system simulation under an impulse-like input. Compares Euler vs Heun (RK2) vsodeint, includes damping-ratio estimation and error plots. -
Neural-Network-Classification-Casting-Defect-Detection
Python code extracted from an academic Jupyter submission. Covers classification, hyperparameter tuning, and casting defect detection using MLPs and CNNs. -
Blasius Boundary-Layer Calculator (Shooting Method)
Numerical solver for the Blasius similarity ODE using a shooting method + root finding, with velocity profiles and boundary-layer metrics (e.g.,f''(0),delta_99, displacement thickness).
- Data-centric engineering: building pipelines around real test/production data (cleaning, feature engineering, uncertainty, model validation) to improve engineering decisions and system performance.
- Packaging future projects into reproducible scripts (less “notebook-style”), with clearer inputs/outputs.
- Python: NumPy, SciPy, Matplotlib, OpenCV, TensorFlow/Keras
- MATLAB: Control System Toolbox, simulation + plotting workflows
- CAD/CFD (when needed): SolidWorks, ANSYS (Fluent/Mechanical)
Alongside uni, I’ve worked in hands-on and client-facing roles (including plumbing/site work and compliance auditing), which has made me comfortable owning tasks end-to-end, communicating clearly and working to standards.