How Indonesia's BRIN Is Using AI to Simulate Ocean Waves in the Lab

Indonesia's national research agency is bringing artificial intelligence to one of the most challenging problems in maritime engineering: testing technology in realistic ocean conditions without ever leaving the laboratory.

Badan Riset dan Inovasi Nasional (BRIN) has developed an AI-based control system for ocean wave simulators that could dramatically reduce the cost and complexity of testing maritime technology. The system uses the Salp Swarm Algorithm (SSA)—a metaheuristic optimization technique inspired by the chain formation of salps in the ocean—to fine-tune PID controllers that coordinate a Stewart platform, a six-degree-of-freedom motion platform capable of replicating complex wave motions.

Why Wave Simulators Matter

Testing maritime technology directly at sea presents enormous practical challenges. Ocean trials require vessels, crew, permits, and weather windows—often costing hundreds of thousands of dollars per deployment. More problematically, sea conditions are unpredictable and uncontrollable, making it difficult to reproduce specific scenarios for comparative testing.

Wave simulators solve this by bringing the ocean to the lab. Using hydraulic or electric actuators mounted on a Stewart platform, researchers can replicate precise wave profiles—height, frequency, direction—thousands of times under identical conditions. This enables systematic testing of ship hull designs, offshore platform stability, navigation systems, and safety equipment.

But getting a simulator to accurately follow real ocean wave patterns is harder than it sounds. The control system must coordinate six actuators simultaneously, compensating for dynamics, delays, and physical constraints—all while maintaining sub-millimeter accuracy.

How SSA Improves the Control System

BRIN's approach combines several technical innovations:

1. Ocean wave models are converted into trajectory paths using inverse kinematics, which maps desired wave profiles onto the physical movement of each platform leg
2. A PID controller ensures the platform follows those trajectories
3. The Salp Swarm Algorithm optimizes the PID controller's gains, searching through parameter space more efficiently than traditional methods

The key breakthrough is using SSA to avoid the local minima that plague other optimization approaches. When compared against Genetic Algorithm (GA) and Particle Swarm Optimization (PSO), the SSA-optimized controller achieved a fitness value 16.8% lower than GA and 8.7% lower than PSO. In practical terms, this means the platform tracks wave profiles with significantly less error—a critical factor when testing equipment that must perform reliably in rough seas.

Optimization Method	Relative Fitness	Performance
Salp Swarm Algorithm (SSA)	Baseline	Best
Particle Swarm Optimization (PSO)	8.7% higher	Good
Genetic Algorithm (GA)	16.8% higher	Weakest

Real-World Applications

The technology has immediate practical uses across multiple maritime domains:

• Ship design: Test hull configurations in simulated storm conditions without building expensive prototypes
• Offshore wind: Validate stability systems for wind platforms before deployment
• Autonomous vessels: Train autonomous surface vessels in realistic wave environments before ever touching saltwater
• Extreme event testing: Simulate rare edge cases or precisely controlled conditions for hundreds of repeated trials—an impossibility with full-scale ocean testing

Perhaps most significantly, the system enables testing that would simply be impossible at sea. Researchers can simulate extreme wave events, rare edge cases, or precisely controlled conditions for hundreds of repeated trials.

Looking Forward

BRIN plans to extend the system with nonlinear control methods capable of handling even more complex dynamic conditions. This could expand applications into areas like wave energy converters, which must optimize energy capture across constantly changing sea states.

The broader implication is clear: AI-driven optimization is making physical simulation systems more accurate, more efficient, and more accessible. What once required massive infrastructure and budgets can now be achieved in laboratory settings—potentially accelerating maritime innovation across Southeast Asia and beyond.