[Paper Review] Intelligent Drone Swarm for Search and Rescue Operations at Sea

Vincenzo Lomonaco, Angelo Trotta+3 more|arXiv (Cornell University)|Nov 13, 2018

UAV Applications and Optimization4 references18 citations

TL;DR

This paper proposes an intelligent drone swarm using autonomous UAVs equipped with AI to enhance search and rescue (SAR) operations in the Mediterranean Sea, addressing migrant crisis challenges. By leveraging self-organizing multi-hop communication, real-time AI-based detection, and distributed coordination, the swarm enables large-area coverage, improved communication bandwidth, and reduced reliance on human-piloted assets, significantly increasing SAR efficiency and scalability in high-risk maritime environments.

ABSTRACT

In recent years, a rising numbers of people arrived in the European Union, traveling across the Mediterranean Sea or overland through Southeast Europe in what has been later named as the European migrant crisis. In the last 5 years, more than 16 thousands people have lost their lives in the Mediterranean sea during the crossing. The United Nations Secretary General Strategy on New Technologies is supporting the use of Artificial Intelligence (AI) and Robotics to accelerate the achievement of the 2030 Sustainable Development Agenda, which includes safe and regular migration processes among the others. In the same spirit, the central idea of this project aims at using AI technology for Search And Rescue (SAR) operations at sea. In particular, we propose an autonomous fleet of self-organizing intelligent drones that would enable the coverage of a broader area, speeding-up the search processes and finally increasing the efficiency and effectiveness of migrants rescue operations.

Motivation & Objective

• Address the urgent need for scalable, cost-effective, and efficient search and rescue (SAR) operations in the Mediterranean Sea, particularly during the migrant crisis.
• Overcome limitations of current SAR systems, including limited flight endurance, restricted communication range, and high operational costs of manned aircraft and ships.
• Enable large-scale, real-time monitoring of vast maritime areas (e.g., 30×30 km²) through coordinated autonomous UAV swarms.
• Improve communication resilience and bandwidth by establishing multi-hop aerial communication networks using line-of-sight (LoS) links between drones.
• Integrate on-board AI for real-time detection and data filtering to minimize downlink traffic and reduce energy consumption.

Proposed method

• Deploy a fleet of fixed-wing, off-the-shelf UAVs equipped with AI inference chips and lightweight detection algorithms to enable on-board processing.
• Implement a self-organizing swarm architecture that dynamically adjusts flight paths and communication parameters based on environmental conditions and mission objectives.
• Establish a multi-hop ad-hoc communication network using aerial-to-aerial LoS links to extend communication range beyond traditional ground-to-air limits.
• Use AI-based object detection (e.g., person-in-water detection) running locally on each UAV to reduce data transmission to the base station.
• Apply model quantization and pruning techniques to reduce computational and energy demands of AI inference on resource-constrained UAV platforms.
• Design a decentralized coordination framework where UAVs autonomously maintain formation, share situational awareness, and adapt to failures or environmental disturbances.

Experimental results

Research questions

• RQ1 Can an autonomous drone swarm effectively cover large maritime SAR areas (e.g., 30×30 km²) beyond the LoS range of a single UAV?
• RQ2 How can AI-enabled on-board processing reduce data transmission load and improve communication efficiency in a drone swarm?
• RQ3 To what extent can a self-organizing multi-hop communication network maintain connectivity and bandwidth in open-sea environments with high mobility?
• RQ4 How does the integration of lightweight AI models on UAVs improve detection accuracy and response time in real-world SAR scenarios?
• RQ5 What are the key technical and operational challenges in deploying such a swarm in hostile maritime conditions, and how can they be mitigated?

Key findings

• The proposed drone swarm architecture enables coverage of large maritime areas (e.g., 30×30 km²) that exceed the operational range and endurance of single UAVs.
• Multi-hop aerial communication links significantly extend communication range and improve bandwidth compared to traditional ground-to-air links.
• On-board AI detection algorithms reduce the amount of data transmitted to the base station, minimizing latency and energy use.
• The use of model quantization and pruning allows efficient AI inference on low-power UAV platforms, supporting real-time processing with limited computational resources.
• The swarm’s decentralized coordination improves resilience to individual UAV failures and environmental disruptions such as wind.
• The system offers a scalable and sustainable alternative to current SAR operations, potentially reducing long-term costs and improving response speed in life-threatening scenarios.