AI-Powered Site Investigation for Renewable Energy Generation

This is how we helped set new sites of renewable energy by detecting materials on the seabed. Learn more.

Computer Vision

Machine Learning

Image Processing

Industry

Energy

Service

Scale your team

About Orsted

∅rsted is a leader in the renewable energy industry.

In partnership with Newlab Blue Energy Studio, their main goal is to investigate and instal new sites for renewable energy generation, improve efficiency in site operations, and optimize power distribution to energy grids.

∅rsted’s team focuses on mapping and detecting material on the seabed using side-scan sonar data. They manually map positions and dimensions of boulders and debris. These objects of interest can get up to 20k+.

The Challenge

To make the scanning process more efficient. The manual and inexact process costs delays in wind farm projects and up to millions of dollars daily.

To validate quickly and efficiently the recommended approach, we had to:

Understand the problem and review of the state-of-the-art methods.
Selection of the ML and IP tools.
Data preparation and labeling.
Development and Training of ML models.
Performance evaluation and KPIs estimation.
Report main findings and recommendations.

The Solution

Automate the process while improving the efficiency and accuracy of seabed material detection using machine learning and image processing.

To detect and measure boulders in SSS images that are noisy and ambiguous is not easy. Failing to do so, impacts on the time and costs of the entire project. An accurate detection is key.