UK-based marine systems provider ASV Global, in partnership with engineering, science and technology consultancy BMT, is set to carry out a research project to improve the safety and reliability of autonomous navigation at sea.
Under the £1.2m Synthetic Imagery training for Machine Vision in Extreme Environments (SIMVEE) project, the partnership will apply deep learning machine vision systems trained with a combination of simulated and real-world data.
The new project will receive some portions of its fund from Innovate UK.
It will primarily work towards enhancing situational awareness to allow unmanned surface vehicles (USVs) to operate in extreme and congested marine environments.
The project will involve the use of ASV’s current, COLREG cognisant, autonomous collision avoidance, and path planning capability.
It will also use BMT’s REMBRANDT ship manoeuvring simulator to train and authenticate ASV’s vision algorithms to detect and classify objects at sea.
The project is expected to result in a better situational awareness solution for both the autonomy on-board and the remote human supervisor.
ASV Global R&D director Richard Daltry said: “This work will provide a significant step in the capability of ASV Global’s ASView autonomous control and navigation system.
“Today we use a remote human supervisor and AIS to classify objects and ensure safe operations.
“The addition of machine vision that detects and classifies objects extends our COLREG compliant autonomous navigation, enabling operations in limited bandwidth with reduced supervisor workload.”
The SIMVEE project aims to train the autonomy system with large quantities of data using the information to be collected by BMT’s REMBRANDT simulator.
This process offers an inexpensive way to collect data and advance the machine learning process.
The SIMVEE project also seeks to enable autonomous surface vehicles (ASVs) to run the same way as traditional manned vessels, as well as launch new use cases and applications with the use of BMT’s Search and Rescue Information System (SARIS).