AQUAFIND: Local Wave Forecasting for Safer Offshore Operations
Offshore activities, such as the installation and maintenance of wind turbines, are highly dependent on wind, waves and rapidly changing weather conditions. Uncertainty about local wave conditions can lead to work at sea being postponed or halted prematurely, even when this later proves unnecessary. More accurate local forecasts can help offshore operators make safer and more efficient decisions about when to start, continue or temporarily suspend operations.
In AQUAFIND (Aerial Quadcopter Units for Aquatic Flow Investigation and Nautical Data), Deltares and its partners are developing a system that helps offshore operations look further ahead. Autonomous drones measure waves, wind and other environmental parameters at sea. By combining these measurements with wave models and artificial intelligence (AI), the system generates a local forecast for the specific location where operations are taking place.
This provides a more accurate picture of the conditions surrounding a vessel or offshore installation, between 30 and 60 minutes in advance. The forecasts can help operators plan activities more safely, reduce unnecessary downtime and deploy vessels, crews and equipment more efficiently.
Growing offshore wind sector increases demand for local forecasts
The expansion of offshore wind energy, ongoing maintenance of existing wind farms, and increasing maritime traffic and other offshore activities are driving demand for safe and efficient offshore operations. Reliable weather and wave forecasts are essential for all of these activities.
Existing monitoring and forecasting methods do not always provide sufficient detail at the exact location of a vessel. As a result, operators often work with wide safety margins. While necessary and understandable, these margins can sometimes be restrictive. Improved local information could extend operational weather windows without compromising safety.
Drones measure waves before they reach the vessel
The concept behind AQUAFIND is not only to measure conditions at the vessel's location, but also further offshore, where incoming waves come from. Swarms of drones operate at some distance from the ongoing activities, for example 10 to 20 kilometres from the vessel, collecting detailed information on waves, wind and other environmental parameters.
These measurements provide an early indication of future conditions. Waves detected further offshore may reach the operational site 30 to 60 minutes later. By integrating drone measurements with wave models and AI, the system creates a local forecast for the next half hour to hour. This forecast can then be incorporated into decision-support tools that help operators decide whether to continue, pause or stop operations.
The initial application focuses on offshore activities in deeper waters, including the installation, inspection and maintenance of offshore wind farms. Nearshore and estuarine environments often involve more complex wave conditions, such as breaking waves or irregular seabed topography. Further development will be required before the technology can be applied effectively in these areas.
From wave measurements to reliable forecasts
Producing reliable local wave forecasts requires close integration of measurements, models, AI and operational expertise. AQUAFIND therefore brings together several partners, each contributing to a different part of the system. SpectX is developing the drones and sensors for measuring waves, wind and other environmental parameters. Delft University of Technology (TU Delft) is investigating how drone swarms can operate safely. Delta-N is developing the AI model and creating a user interface that combines measurements and forecasts. TNO is developing decision-support tools for offshore operations and is responsible for testing and validating the technology.
Deltares connects these components through its expertise in wave processes, modelling and operational forecasting. We are investigating how wave fields measured by drones can be translated into reliable forecasts at the vessel location. To support this work, we use results from the SWAN wave model to build an extensive training dataset for the AI model. We also compare the new forecasts with existing operational wave predictions and provide experimental facilities for testing sensors and prototypes against in-situ wave measurements.
Innovation through the integration of drones, wave models and AI
The innovation within AQUAFIND lies in the way autonomous drone swarms, real-time wave measurements, physical wave models and AI are combined. Together, these technologies make it possible not only to describe current sea conditions but also to predict how conditions will evolve in the short term at the location where operations are taking place.
The knowledge developed within the project could also be valuable beyond this initial application. Potential future applications include drone-based wave monitoring, AI emulators for wave models and new approaches for using autonomous drones in offshore measurement campaigns.
Next step: testing and validating the prototype
AQUAFIND is a collaboration between Deltares, SpectX, Delta-N, TU Delft and TNO. Several offshore companies are also involved through an advisory group to ensure that the technology is aligned with operational practice at sea.
AQUAFIND is a MOOI project within the Top Sector Energy programme. The project runs from November 2024 until February 2029.
Researchers are currently focusing on the development of both the AI model and the drones. The next major step will be testing and validating the prototype under offshore conditions. This will provide insight into how the system performs at sea and what is required for acceptance and adoption by the offshore sector.
