Case study / DATA / OCEAN
WaveRadar
Internship completedA real-time ocean visualisation project developed at OEG during the 2026 EEEGR Coastal Energy Internship Programme, turning wave-spectrum data into a computational 3D surface.
This was my primary project during a four-week, 20-day paid internship at OEG in Lowestoft in August 2026, completed through the EEEGR Coastal Energy Internship Programme.
Industry context
The official project brief was Digital Ocean: Turning Wave Radar Data into 3D Visualisation. My objective was to investigate and implement a way to transform wave-spectrum data into a real-time 3D ocean representation.
This was an internship visualisation project, distinct from OEG’s commercial WaveRadar systems. I am not claiming development of the company’s production product or production software.
From spectrum to surface
Wave spectra describe how energy is distributed across frequency and direction. The implementation constructs a frequency-domain representation and evolves it over time.
Inverse FFT operations turn that representation into spatial fields for surface height and horizontal displacement. A further pass calculates the normals used to render the surface. Together, these steps turn numerical wave information into a moving 3D ocean.
Computational analysis
The workload study is based on what the implemented code performs: spectral synthesis, FFT passes, displacement and normal calculations. It is not an abstract calculation detached from a working implementation.
The accompanying calculator varies resolution and frame rate to explore estimated operation and thread counts. These are estimates derived from the algorithm and code, not measured GPU timings, benchmark results or guaranteed frame rates.
Explore the simulation calculations →
Outcome and limits
The placement brought together practical implementation, real-time visualisation and computational analysis within a defined engineering-industry project. I presented the work and results at the end of the internship and gained experience of professional engineering and software workflows.
The value of the visualisation is in making the numerical model inspectable. A convincing moving surface does not, on its own, establish physical validation, and an operation count does not establish hardware performance.