Background

RIGOCAL (which stands for "RIGOrous CALculation") provides civil engineering, surveying and marine mammal observation services for energy, construction and offshore industries.

The guideline for offshore piling in the UK Continental Shelf, set by the Joint Nature Conservation Committee (JNCC), requires that JNCC certified Marine Mammals Observers (MMO) should monitor the presence of marine mammals within a mitigation zone around the source of noise during pile driving for offshore wind construction.  The most popular way to do this is by using binoculars with the data being recorded manually by MMOs. This method faces criticism because it does not have a wide view of the area (360 degrees) and it can only be done in good visibility (daylight) with good sea conditions.

RIGOCAL is architecting the best way to use infrared (IR) cameras to detect marine mammals when they approach the sea surface in the vicinity of offshore structures under construction. Their novel concept combines machine learning with Infrared and High Definition videos to automatically detect and classify marine mammals.

 

Challenge

The company was looking to do a feasibility study into the development of an algorithm for the automatic extraction of marine mammals (seals, dolphins, whales, porpoises) in video images from thermally detected radiations (infrared camera) and high definition images (RGB camera).

 

Solution

Scottish Enterprise referred RIGOCAL on to Interface, who then put the company in contact with the Edinburgh Parallel Computing Centre (EPCC) at the University of Edinburgh as they have expertise in advanced data analytics and image recognition capability based on the utilisation of machine learning technologies.  In collaborating on this project, EPCC were able to make use of its specialist facilities - namely its High-Performance Computing and Data Infrastructure located within the University.

The feasibility study was funded by a Scottish Funding Council Innovation Voucher.

 

Benefits

Company Benefits

EPCC closed a gap in RIGOCAL’s technical and resource capabilities for developing a machine-learning algorithm to analyse data in real-time. This innovation will change the way traditional marine mammal observation is conducted providing more accurate and consistent service at a lower cost.

The RIGOCAL approach allows a 360-degree view of the area. This approach is also innovative in that marine mammal detection will be made possible during the night and in poor visibility because of the advanced IR sensor. This will allow pile driving to start even in low visibility, especially night time, which will enable the offshore piling construction companies to operate 24/7.

As a result, RIGOCAL’s new service will contribute to reducing total operation hours, therefore cost.  Cost reduction during the construction phase will lead to total cost reduction in offshore wind development, helping the industry to achieve the UK government’s goal of driving down the levelised cost of electricity.

The technology applied for this specific environmental issue will also open up new opportunities, not only from additional environmental perspectives - such as protecting birds against collision with the blades of wind turbines, and leak detection from oil pipelines - it will also save human life in the sea by allowing automatic detection of castaways.

Dr. Alessandro Bedin

The partnership of RIGOCAL with EPCC for this feasibility study has exceeded RIGOCAL’s expectations and we are looking forward to continuing our collaboration with them for the development of the algorithm itself.”  

Dr Alessandro Bedin, Managing Director, RIGOCAL Engineering Ltd

Academic Benefits

EPCC will benefit from:

(1) increased capability in the application of advanced data analytics and machine learning,

(2) an opportunity to grow its target market across the marine/energy space, and

(3) strengthening its position as a sustainable centre collaborating with industry.

Terry Sloan

“This was a technically challenging but satisfying project in which we worked closely with RIGOCAL to apply existing deep learning models to a new area. Our collaboration has produced a more efficient service for RIGOCAL’s customers and we hope it will contribute to the conservation of marine mammals."

Terry Sloan, EPCC Project Manager, The University of Edinburgh