Key Takeaways:

• SAIC has proven that collaborative autonomy works: At Talisman Sabre 2025, we deployed four-vehicle UUV teams that executed complex missions with zero human operators in the loop. The vehicles coordinated autonomously using minimal underwater communications, while users simply monitored the output.
• Our simulation-to-deployment pipeline significantly compresses timelines: Using AI-optimized simulation, surrogate testing and digital twins, we reduced in-water testing time by 50%.
• We’re investing ahead of demand: Rather than waiting for government contracts to fund development, SAIC is making significant internal investments in AI/ML and technology partnerships to deliver mature collaborative autonomy solutions.

The U.S. Navy faces a proliferating undersea threat environment unlike anything in its history. Across the Indo-Pacific region, North Sea and Red Sea, U.S. partners have seen a rise in underwater activities and probes. But the potential threats are not monolithic. Low-cost commercial drones modified for military use represent one challenge; advanced autonomous systems with sophisticated AI pose another entirely.

The establishment of the Navy’s Portfolio Acquisition Executive for Robotic and Autonomous Systems (PAE RAS) signals the urgency. Fleets of autonomous vehicles will be essential to maintaining undersea dominance. Yet acquiring systems is only part of the answer. 

Two fundamental problems make traditional approaches unviable in deploying UUVs. The first problem is the severe constraints of communication underwater. GPS signals don’t penetrate water, and acoustic communications are slower and have lower bandwidth compared to their surface-level counterparts. They degrade rapidly with distance, interference and adversary jamming. Consequently, it’s impossible to remotely and reliably control multiple vehicles underwater. 

The second problem is that human operators can’t scale. Traditionally, each unmanned underwater vehicle (UUV) requires constant human input. At 10 vehicles, operator burden becomes unmanageable. At 50 or 100, the scale that future operations require, it would be nearly impossible. 

Collaborative autonomy: The ‘playbook’ approach

Collaborative autonomy reimagines unmanned systems as autonomous teammates capable of independent action within mission parameters. UUVs need to function like a football team, in which every member has memorized the playbook. 

For fans watching a game, such coordination can be breathtaking. When the quarterback calls a play or the team perceives that the play is broken and a new one is needed on the fly, every player on the field knows their role. A wide receiver running a route doesn’t need instructions when the ball arrives; he reacts as trained. There is little to no verbal communication during the play itself. That’s essential because crowd noise in a stadium results in “communication blackouts” very similar to conditions undersea.

A “playbook” approach is precisely what we’re building for the underwater domain. In practice, this means four UUVs patrolling in formation can autonomously respond when one detects an adversary vehicle. The team determines who intercepts based on geometry and individual capabilities. Maybe the detecting vehicle is best positioned; perhaps another carries the right payload. The rest regroup to maintain coverage. The team then reports back to command (“Here’s what we found. Awaiting guidance”).

Instead of one expensive platform trying to do everything, systems that are smaller, cheaper and more specialized work together. This approach reduces both cost and communications strain by designating a single “spokesperson” vehicle rather than having every system report individually.

From simulation to operational reality: Talisman Sabre 2025

Theory became practice during Talisman Sabre 2025, when SAIC participated in the Office of the Secretary of Defense’s Resilient and Autonomous AI Technology (RAAIT) trials. This large-scale U.S.-Australia exercise in the Indo-Pacific provided what university laboratories cannot: a real operational environment with surface interference, seafloor debris and the unpredictable conditions that stress-test every assumption.

SAIC deployed four vehicles with no human operators in the loop. Users simply watched the output as vehicles executed their mission using very limited underwater communication. 

Our development pipeline began with more than 20,000 AI/ML-optimized simulations, progressed through surrogate testing on ground robots equipped with LIDAR, advanced to digital twin environments and culminated in live water operations. This simulation-first approach reduced in-water testing time by more than 50%, which is critical for accelerating capability delivery.

Results validated the methodology, with strong simulation accuracy translating effectively to real-world operations. Gaps taught valuable lessons—namely, that algorithms need to emphasize prediction over real-time coordination and that plays must be able to adapt based on actual acoustic conditions. Each iteration made the system smarter, demonstrating the feedback loop that continuous operational learning enables.

Proven capability—ready to scale

Existing hardware solutions can scale to the numbers needed. What’s been missing is scalable coordination—so diverse systems from multiple manufacturers can work together within a collaborative framework. 

That’s what SAIC has invested in building and has now demonstrated.

We have made significant internal investments because we view this capability as essential to national security. We’re investing in our own technology and in technology partnerships to bring mature solutions to government customers. When rapid capability organizations are ready to move, we’ll already be in position to help.

What’s needed now is continued investment in AI/ML to expand playbooks across more scenarios, persistent experimentation in realistic environments, scaling from four-vehicle teams to those of 100 or more, and deeper integration with allies through frameworks like AUKUS.

The imperative for action

Talisman Sabre proved that collaborative autonomy is technically feasible. The question now is how quickly the Navy can field these capabilities at scale. SAIC has moved from concept to operational demonstration, providing a simulation-to-deployment pipeline that compresses timelines while maintaining accuracy. As a platform-agnostic integrator, we’re positioned to help the Navy achieve collaborative autonomy across diverse vehicle types and operational environments. 

The technology is proven and the capability is ready. Let’s scale it together.

Explore how SAIC is advancing collaborative autonomy to help the U.S. Navy counter undersea threats—scaling unmanned systems through AI-driven coordination, resilient design and rapid transition from simulation to operations.

Learn more about our Undersea Capabilities