Multi-modal coordinated sensing for spatially-aware marine monitoring

Accurate location monitoring of marine life is important for sustainable aquaculture, fisheries, and ecosystem management. Current aerial and underwater drones face limits in accuracy and scale. Aerial drones struggle with depth and glare; underwater drones have a short range and a narrow field of view. Multi-drone systems scale, but coordination over water is hard because GPS is unreliable and noise is high. This paper presents a Coordinated Multi Drone (CMD) architecture that integrates an aerial drone and a surface drone for multimodal sensing. The aerial drone detects surface targets, and the surface drone, with an underwater camera, performs near-field classification. A kernel-based fusion algorithm merges observations in a shared Star Marker Positioning (SMP) frame for precise localization, achieving 92% spatial alignment. Seaweed is the test case, but the approach generalizes to other marine biomass. Field trials achieved 95% within 4 m, with 0.02-0.12 m localization error and a 25% improvement in underwater detection. CMD provides a sensing framework for spatially aware marine monitoring using coordinated, multimodal platforms.