Heterogeneous Multirobot Team: Maritime Inspection and Intervention in Global Navigation Satellite System-Denied Scenarios
Lovro Markovi ́c, Robert Milijaˇs, Dario Stuhne, Matko Orsag, Stjepan Bogdan,, and Nikola Miˇskovi ́c
AI summary
Problem
Autonomous maritime operations heavily depend on GNSS, which is often unavailable or unreliable in critical coastal and open-sea scenarios. Current sea-aerial robotic systems lack integrated, robust solutions for precise navigation, docking, and cooperative manipulation without satellite signals.
Approach
The team engineered a marsupial configuration where an ASV carries and charges UAVs, fusing LoRa-based ranging, DVL/IMU navigation, and vision-assisted tracking to enable precise localization and coordination in featureless, satellite-denied waters.
Key results
- First-place win at the MBZIRC 2023 Maritime Grand Challenge
- Robust GNSS-denied localization via LoRa and DVL/IMU fusion
- Autonomous docking and cooperative object retrieval with a robotic arm
- Validated real-world ASV-UAV collaboration in adverse maritime conditions
Why it matters
Enables reliable autonomous maritime inspection, search-and-rescue, and intervention missions in environments where traditional satellite navigation fails.
Abstract
Autonomous robotic systems play an increasingly important role in overcoming complex maritime challenges where the environment requires high adaptability, robustness, and precision. Marsupial robotic systems, combining aerial systems and surface vessels, offer a powerful solution for coordinated missions in GNSS-denied scenarios, leveraging the unique capabilities of each robot to form a versatile team. This paper presents a system integrating advanced sensor fusion, localization, mission planning, and communication to enable seamless collaboration between Uncrewed Aerial Vehicles (UAVs) and Autonomous Surface Vehicles (ASVs). The focus is on optimizing their interaction to efficiently locate, retrieve, and transport objects from a target vessel in an adverse sea environment. This paper provides a complete system overview, including the hardware setup, algorithms used, implementation details, and field trial results. The results analysis demonstrates the advanced capabilities of the developed ASV-UAV system in complex, real-world maritime operations, underscoring its potential for coordinated multi-robot tasks in environments where traditional navigation systems are unavailable. Notably, this system was developed by the team that secured first place in the Mohamed Bin Zayed International Robotic Challenge (MBZIRC) competition, proving its excellence in challenging conditions.