A Conceptual Framework for a Human-Supervised Dual-UUV System for Autonomous Maritime Surveillance
Context
Traditional maritime surveillance systems lack operational flexibility, suffering from high acoustic signal attenuation and wave noise that degrades anomaly tracking. Identifying and classifying targets in real-time requires cooperative multi-vehicle systems to balance cover and sensor footprint.
Research Problem
Traditional maritime surveillance systems often struggle to balance stealthy detection with active response, especially in environments requiring strict ethical oversight. Existing autonomous solutions often lack the "Human-in-the-loop" gating necessary for non-lethal compliance.
Proposed System Architecture
Designed a multi-agent cooperative UUV surveillance systems architecture. The model divides operational roles between a passive, silent 'Scout' UUV (for initial acoustic anomaly detection) and an active, sonar-equipped 'Inspector' UUV (for optical and high-frequency sonar validation), coordinating states over dynamic acoustic communication links.
Core Technical Areas
- â– Autonomous Systems
- â– UUV Research
- â– Maritime Surveillance
- â– Multi-Agent Systems
- â– Sensor Fusion (EKF)
- â– Signal Processing
- â– Human-Supervised Autonomy
- â– Systems Architecture
Key Contributions
- ✓Designed a multi-agent systems architecture dividing scout-detection and inspector-validation roles between specialized UUV classes.
- ✓Formulated a cooperative target localization model utilizing passive acoustic arrays and range-only tracking filters.
- ✓Integrated a Human-Supervised cryptographic override schema, establishing safety guardrails for autonomous mission plans.
- ✓Developed a sensor fusion mathematical model employing Extended Kalman Filtering (EKF) to merge DVL (Doppler Velocity Log) and INS (Inertial Navigation System) telemetry, reducing cumulative positional drift in simulated states.
[!] Limitations & Ethical Considerations
This is a purely conceptual systems design and mathematical modeling framework. No physical submarine vehicle was manufactured, and the system is strictly limited to simulated, human-supervised surveillance modeling.
What I Learned
- Extended Kalman Filtering (EKF) sensor fusion algorithms and error state propagation.
- Acoustic signal propagation under water and communication channel limitations.
- Ethical design patterns for autonomous systems and human-in-the-loop control overrides.
Read the Paper
Full theoretical analysis, localization math, and communication protocols.
System Architecture & Framework