Autonomous underwater vehicles (AUVs) are increasingly used for subsea operations, for example to perform pipeline inspection, mapping of the marine environment, and monitoring the sea temperature, salinity, and ocean currents. Similarly to what is happening for self-driving cars, AUVs are becoming more and more autonomous, and capable to take decisions by themselves. However, when deployed in a multitude, they are currently unable to take advantage of being more than one, and rather operate by themselves as they were alone.
Being unable to collaborate is a clear drawback: cooperating AUVs may perform more complex automated marine operations, for example performing sea inspection tasks in shorter time and with better results, or achieving results that a single AUV may not, such as transporting a pipe.
This project aims at enabling several AUVs to execute tasks together and perform collaborative underwater marine operations, and through this enable more efficient & reliable monitoring and utilization of sea resources & infrastructures.
To reach this, the strategy will be making these vehicles capable of: 1) understanding which information they should exchange with the peers while operating, 2) physically exchanging this data through underwater communications, plus 3) understanding, from the information arriving from the peers, how the individual should adapt its own behavior so that it is not "selfish" anymore, but rather "helping" the others.
The planned results aim thus at transforming sets of independently operating AUVs into fleets of collaborators exchanging messages and operating collaboratively, advancing thus Norway's marine technology and the capability of developing products for the underwater realm.
Autonomous underwater vehicles (AUVs) are increasingly used for subsea operations like pipeline inspection, mapping and monitoring. Their level of autonomy is increasing, but they are still unable to perform cooperative tasks when deployed in a multitude. If collaborating, AUVs can perform automated marine operations such as accomplish sea inspection tasks in shorter time and with better results, plus achieve results that a single AUV may not, such as transporting a pipe.
Our research project will develop algorithms that transform sets of independently operating AUVs into fleets of collaborators exchanging information, deciding collectively, and operating synergically. To do so the project will generate foundational mathematical advancements about distributed control, distributed estimation & optimization, and adaptive acoustic communications algorithms. On top of this, we will merge the fields above into a holistic and comprehensive knowledge for enhancing the autonomy and functionality of teams of underwater vehicles.
These advancements will help Norway's marine technology companies maintain and even improve their world leading positions and enhance the possibilities of developing products for the inspection, monitoring and exploration of underwater structures, or for environmental applications such as monitoring of temperature, salinity, and ocean currents.
The project will be integrated in the Centre of Excellence on Autonomous Marine Operations and Systems at NTNU and will strengthen the cooperation in marine robotics and autonomy with Equinor, DNV GL, and the large group of national research and industry collaborators surrounding AMOS (such as the spin-off company Eelume and Kongsberg Maritime). Close collaborations will be also with the European project OASYS, developing micro underwater gliders for collecting oceanographic data.