Using SOTA Generative Models to predict Remaining Useful Life of Wind turbine Gearboxes
For my final; year research topic I was particularly interested in how machine learning can be used as a tool in the space of maintenance engineering. So under the supervision and guidance of Dr. Wilke, I conducted research and expirements on using machine learning models to predict the remaining useful life of a wind turbine gearbox
-------------------------------------------------------------------------------------------------------------------------------------------
Human-powered aircraft - Propeller Design
For my final year design project. I got my inspiration from reading on the Kramer Price. The Kramer Prize is an international competition whereby engineers and designers all around the world to try build and fly a human powered aircraft. This being such a particularly interesting project i mainly focused on the propeller for a human powered aircraft. Key design consideration such as power to weigh ratios, ease of assembly and production costs had to be thought of when designing the propeller. The document below runs through my research, design process, final design and specifications of the propeller.
Gearbox _ Design
This gearbox design wasn't for a particular project. It was just a design i replicated from a fellow Engineer
Mabone - Lights
This was a design of mine for modern lights using low cost material and energy saving LED strips
Blompot - Fowerplot
This was a particularly interesting project for me. I created this design for a pot plant. Manufacturing it was the tricky part as I was using materials such as concrete (for main body) and copper for the legs. After many failed attempts and bags of cement later the final product was produced.
Vibration Sensing System
For my final year research project. I was focusing on using the vibration data of a gearbox system to build a SOTA generative model to predict the remaining useful life of a gearbox. This required me to build the sensor jig you see in the images to the left. I utilized an Arduino-based microprocessor to connect an ADXL335 accelerometer and record vibration readings that had a sensitivity of +/-16G sensitivity in the X,Y and Z axis..