Exploiting Simulation to Train Underwater Uncrewed Vehicle Swarms: Insights from Talisman Sabre 2025

As adversaries increasingly deploy unmanned underwater vehicles (UUVs) to disrupt maritime operations, the need for scalable, autonomous and collaborative countermeasures is urgent. Future naval dominance will depend not just on superior hardware but on intelligent systems that can act independently, coordinate seamlessly and adapt in real time. It will be especially critical in communication-denied environments. What was once considered a futuristic set of capabilities that would take months to iterate at very high cost has quickly become a deployable undersea warfare sensor and/or weapon system that can be built to order in a matter of days or weeks that is both cost effective and rapidly deployable.

During Talisman Sabre 2025 (TS-25), a joint U.S., Australian and allied military exercise, SAIC, a leading mission integrator deploying the next generation of capabilities for the Navy today and affiliates explored how simulation, artificial intelligence and machine learning could be used to train swarms of UUVs to detect and track adversaries, allocate tasks and operate autonomously all with minimal reliance on centralized command. To achieve this a four-unit UUV swarm was deployed to patrol an area using sonar and acoustic communication systems. Each vehicle executed pre-trained “plays” or algorithmic behaviors that mimic tactical maneuvers, allowing the swarm to respond quickly and cooperatively without external input.

To test potential behavior before deployment of the UUVs, over 20,000 high-fidelity simulations were conducted using a modified version of the Advanced Framework for Simulation Integration and Modeling (AFSIM). The training data informed a neural network capable of weighing tactical variables such as battery life, sonar range and positioning to assign pursuit tasks intelligently, and validation accuracies exceeded 97%. The team used digital twins and Robot Operating System-based surrogates to test AI logic on land using LiDAR, significantly reducing reliance on live testing. This cut development time by more than half while preserving mission logic.

In-water trials were then conducted at depths of 1-10 meters, where the UUVs used Forward-Looking Sonar (FLS) and Doppler Velocity Logs (DVL) for navigation and detection. The vehicles successfully tracked adversaries and adjusted behaviors despite communication delays and environmental noise. Predictive modeling was added to improve response reliability.

Each mission generated datasets tracking logs, sonar returns and message sequences all that fed back into the simulation, forming a continuous feedback loop. This process allows for real-world performance to shape future training, making the system smarter with every iteration.

Designed for scalability and adaptability, the swarm architecture supports low-bandwidth, delay-tolerant communication, modular AI integration and the ability to add new behaviors or sensors quickly. This helps create reliable, low-cost and high-value missions. Future enhancements include heterogeneous teaming where vehicles play distinct roles like scouts or interceptors. What once required months of coding and testing can now be accomplished in weeks trained in simulation, validated on land and deployed live. The result is a resilient, intelligent swarm system capable of executing missions in dynamic, noisy and degraded environments.

This initiative will be presented at I/ITSEC 2025 as a model for rapidly fielding autonomous maritime capabilities. It marks a critical shift from static tools to adaptive, learning-enabled systems—scalable, deployable and mission-ready. In an era of rising undersea threats, this framework offers both technological edge and operational superiority.

Together with the power of mission integration, new commercial capabilities, and ready-now, rapidly deployable, and future-focused technologies, the SAIC team is working with a diverse ecosystem of solution providers to enable the future of undersea warfare. Read the paper that will be presented at ITSEC and learn more here.

×