Data Ingestion and Processing - The system shall be capable of ingesting data from maritime sensors, autonomous vessels, IoT devices, and environmental monitoring systems. - Data shall include but is not limited to vessel position, speed, heading, and telemetry. - Environmental data such as water temperature, wind speed, wave height, and obstacles shall also be available. Ingestion Protocols - The system shall support data ingestion via MQTT, Kafka, WebSocket, and REST APIs. Data Handling - The system shall normalize and clean incoming data streams to ensure consistency across all simulation scenarios. Integration with Existing Models - The TAP shall provide AI-driven simulations of maritime scenarios. The AI shall also predict vessel behaviors, optimize collision avoidance strategies, and simulate environmental interactions. Integration with Existing Models - Bellhop 2D/3D Models: Simulate sound propagation and underwater acoustic scenarios. - NGOFS2/HYCOM Models: Simulate Gulf of Mexico conditions, including hydrodynamic, current, and wave models. - USM’s Hi-Res Coastal Models: For high-resolution coastal and port simulations. - AI Models: Shall support ONNX-based models for performance optimization on the target hardware. - Simulation Tools: These tools shall support integration with tools for real-time 3D vessel behavior modeling and environmental simulation. Edge Computing Deployment - The TAP shall support deployment on high- performance laptops/desktops allowing for field-based use. Some features may be cloud based. - Laptop Specifications: High-end laptops with dedicated GPUs, supporting both Windows and Linux shall be provided. - On-Device AI Inference: Only ONNX and TinyML shall be used to optimize AI model inference on devices with limited resources. - Scalability: The system shall be capable of scaling from local deployment on a single laptop to a cloud- based environment for more extensive simulations.
