{"id":2129,"date":"2025-08-19T10:50:15","date_gmt":"2025-08-19T10:50:15","guid":{"rendered":"https:\/\/oasees-project.eu\/?p=2129"},"modified":"2025-08-19T10:51:06","modified_gmt":"2025-08-19T10:51:06","slug":"oasees-swarmaware-srds-conference-with-3-publications","status":"publish","type":"post","link":"https:\/\/oasees-project.eu\/?p=2129","title":{"rendered":"OASEES @ SwarmAware – SRDS conference with 3 publications – pre-final versions available here"},"content":{"rendered":"\n
\"\"<\/figure>\n\n\n\n


OASEES is excited to participate in the SwarmAware conference with three publications that span foundational design requirements, autonomous aerial swarms, and edge intelligence for renewable energy. Together, these papers showcase our commitment to building scalable, trustworthy, and field-ready swarm systems.<\/p>\n\n\n\n

Designing Swarm-based Decentralised Systems: Requirements for Performance and Scalability<\/strong><\/h2>\n\n\n\n

<\/p>\n\n\n\n

This paper distills the architectural and operational requirements that make large-scale, decentralised swarms both performant and resilient. It frames core design decisions\u2014communication topologies, coordination mechanisms, consensus choices, and task allocation\u2014against measurable KPIs such as latency, throughput, convergence time, and fault tolerance. The result is a practitioner-ready checklist with reference patterns that help teams reason about trade-offs (e.g., decentralisation vs. overhead, local autonomy vs. global optimality) before they commit to a specific stack.<\/p>\n\n\n\n

A Distributed UAV Analytics Framework for DAO-based Swarm Systems<\/strong><\/h2>\n\n\n\n

<\/p>\n\n\n\n

Here we introduce a full analytics pipeline for UAV swarms governed by a DAO. The framework combines on-board edge processing with cooperative model sharing and on-chain coordination, enabling fleets to adapt mission plans while preserving data sovereignty. It outlines how sensing, inference, and decision logic are partitioned across drones and ground infrastructure, and shows how DAO policies can enforce fairness, reputation, and resource budgeting\u2014paving the way for auditable, human-in-the-loop autonomy at scale.<\/p>\n\n\n\n

Federated Learning at the Edge for Wind Turbine Predictive Maintenance<\/strong><\/h2>\n\n\n\n

This paper demonstrates how federated learning (FL) can unlock predictive maintenance for geographically dispersed wind farms without centralising raw data. By training models directly at the edge\u2014on turbine controllers or substation gateways\u2014the approach reduces bandwidth use, respects privacy constraints, and adapts to local operating conditions. The study highlights gains in early fault detection and model robustness under non-IID data, offering a clear path from lab prototypes to production-grade condition monitoring.<\/p>\n\n\n\n

list of papers – pre-final versions:<\/p>\n\n\n\n

\n