BLUESIM is an R&D project focused on the development of a high-fidelity synthetic simulation platform for the training and validation of autonomous unmanned underwater vehicles (UUVs).
The initiative investigates a hyper-realistic virtual environment capable of generating millions of hours of traceable and reproducible synthetic data to train Deep Learning and Deep Reinforcement Learning algorithms. Built on advanced graphics engines and accurate physical models, BLUESIM integrates marine environment simulation, vehicle dynamic models, and synthetic sensors, enabling the validation of autonomous technologies in safe and controlled conditions before real-world deployment.
The ultimate goal of the project is to mature the technology up to TRL3 (Proof of Concept), demonstrating the feasibility of a reliable synthetic training environment for autonomous navigation in complex and safety-critical maritime scenarios.
Celestia TTI’s Contribution
Within the project, Celestia TTI has led the design and development of the synthetic GNSS receiver module and the node-to-node communications system integrated into the platform.
Our contribution included defining the architecture of the simulated satellite positioning system, modelling GNSS receiver processing, and ensuring seamless integration within the simulation environment. In addition, Celestia TTI developed the node-to-node communication framework between multiple UUVs, enabling real-time sharing of their estimated positions as well as GNSS signal quality indicators. This capability enhances cooperative behavior and situational awareness among autonomous platforms operating in the simulated maritime environment.
Through this participation, Celestia TTI further strengthens its expertise in satellite navigation technologies, advanced communications systems, and synthetic simulation environments for autonomous maritime applications.
