SMD
The company
SMD is a trusted subsea vehicle developer with over 50 years of experience. They are renowned for their reliable systems and innovative solutions. The company specialises in creating subsea vehicles such as remotely operated vehicles (ROVs), ploughs, and trenchers that perform heavy lifting and intricate tasks beneath the ocean's surface. These vehicles are crucial for laying cables that provide internet and power, as well as pipelines that transport gas.
In their continuous pursuit of excellence, SMD partnered with the National Innovation Centre for Data (NICD) to develop an advanced object and anomaly detection module for their underwater vehicles. This project aimed to enhance the efficiency and user-friendliness of SMD's subsea vehicles, assisting pilots in navigating challenging underwater environments.
The challenge
The underwater environment presents unique challenges, such as poor visibility caused by murky water and very limited lighting. Pilots often have to navigate through conditions akin to driving through fog or heavy rain, making it difficult to identify and react to objects in their path. SMD needed a solution that could enhance the visual data from their vehicles' cameras, making it easier for pilots to perform their tasks efficiently and safely.
“At the moment, the video from the cameras is just shown to the pilots, and they need to make any inferences based on what they see. Anything we can do to make that easier for them is only going to be a good thing.”
Andrew Ambrose-Thurman, Lead Engineer Autonomous Systems, SMD
The collaboration with NICD
The main focuses of the project were to:
- undertake a literature search to review viable computer vision models
- acquire appropriate data for development and testing
- develop and test multiple computer vision models
- develop a system based on the output to the research outlined above
“The field of computer vision, detection and segmentation, is undergoing a technological revolution. This was the perfect moment for SMD to pursue a project of this nature. It was my great pleasure to contribute to this initiative and support them in delivering something I would have considered impossible just six months earlier.”
Dr Matt Edwards, Former Senior Data Scientist, National Innovation Centre for Data
The results
The project culminated in a successful demonstration of how the latest computer vision systems can be used to successfully categorise underwater Infrastructure such as partially exposed pipes. The system required no training and had the potential to detect other underwater infrastructure.
The demonstration provided clear evidence of the module's ability to enhance the visual data from the vehicles' cameras, making it easier for pilots to identify and react to objects underwater. This capability is expected to significantly improve the efficiency and safety of SMD's subsea operations.
“It was great fun working on this project with SMD. The project was using the latest cutting-edge advances in computer vision in an underwater environment, which made it an interesting and challenging problem to tackle.”
Dr Georgia Atkinson, Data Scientist, National Innovation Centre for Data
“When we first came to NICD, we had some ideas about what we wanted to achieve, but we didn't have any ideas about how we were going to achieve it. It has been very useful being able to work with the people at NICD who have much greater expertise in this area.”
Andrew Ambrose-Thurman, Lead Engineer Autonomous Systems, SMD
To discover more about SMD, visit their website.
You can read more of our case studies and sign up to our newsletter to keep up to date with our latest news, events and developments.