Building Reliable Swarm Systems at the Edge: Insights from the SwarmAware Workshop at SRDS 2025

The P2CODE project took an active role at the 44th International Symposium on Reliable Distributed Systems (SRDS 2025), held from 29 September to 2 October 2025 in Porto, Portugal, by co-organizing the SwarmAware 2025 Workshop on Trustworthy, Decentralized Computing at the Intelligent Edge. The workshop was organized in collaboration with the TaRDIS and OASEES projects and brought together researchers exploring how large, distributed systems of autonomous devices can coordinate, adapt, and operate safely at scale.

The organizing committee was led by Vrettos Moulos (ARHS Group, P2CODE), Carla Ferreira (NOVA University Lisbon, TaRDIS), and Akis Kourtis (NCSR Demokritos, OASEES), reflecting the close cooperation between the three Horizon Europe initiatives focused on secure, intelligent, and dependable distributed systems.

Exploring the Capabilities and Limits of Swarm-Based Systems

SwarmAware provided a forum to examine the emergence of swarm intelligence in edge-to-cloud systems, where large numbers of autonomous agents—from IoT sensors to drones and mobile robots—need to collaborate under uncertain real-world conditions. Discussions throughout the day highlighted both the promise of distributed decision-making and the engineering challenges it introduces in reliability, coordination, and security.

In several talks, researchers demonstrated how swarm-intelligence approaches can improve throughput, scalability, and distributed data processing in dynamic environments. At the same time, others focused on the mechanisms needed to ensure sandboxed execution, model integrity, and data protection, reminding that real-world deployments require both adaptability and restraint. The result was a balanced discussion that linked performance innovation with the need for provable trust and operational safeguards.

A notable contribution from P2CODE partners addressed exactly this intersection. The paper “A Scalable Swarm Intelligence Algorithm for Autonomous UAV Search and Rescue Operations”, authored by Javier Saez-Perez, Julio Diez-Tomillo, Pablo Benlloch-Caballero, Pablo Salva-Garcia and Qi Wang (University of the West of Scotland), showcased a scalable algorithm enabling drone fleets to coordinate during search-and-rescue missions. The work demonstrated how swarm-based approaches can improve coverage efficiency and adaptability in dynamic, high-stakes scenarios—an area where traditional centralized control systems often face limitations.

Across the workshop, several presentations focused on decentralized UAV coordination, edge-based data analytics for drone formations, and collaborative perception architectures. Complementary sessions examined federated learning across distributed nodes, privacy-preserving edge intelligence, and graph-based optimization for dynamic task allocation in cloud-to-edge environments.

Security, Trust, and Dependability as Central Themes

SwarmAware also underscored the importance of trust guarantees in decentralized systems. Participants discussed approaches to ensure that autonomous nodes follow intended behaviors, that models remain interpretable, and that data exchanges can be validated and audited. These issues are deeply aligned with P2CODE’s work on zero-trust communication, attestation, and verifiable data provenance in multi-domain computing environments.

The discussions illustrated that while swarm-based systems promise scalability, adaptability, and resilience, their adoption in critical sectors depends on ensuring reliability, explainability, and robust operational safeguards. Participants emphasized that swarm-based and decentralized intelligence are not research abstractions, but technologies emerging in sectors such as:

  • Search-and-rescue and public protection using coordinated UAV and ground robotics
  • Industrial inspection and predictive maintenance leveraging federated learning at the edge
  • Energy and mobility infrastructures requiring resilient sensing and decision support
  • Smart factories combining autonomy with human oversight

Toward a Shared Research Roadmap

By being held under the umbrella of SRDS, one of the primary reference venues for reliable distributed computing, the SwarmAware workshop highlighted the need for cross-disciplinary alignment between distributed systems theory, edge computing, robotics, and AI safety research. The organizers expressed interest in expanding the workshop series to build a sustained research community focused on trustworthy swarm intelligence at scale.

For P2CODE, SwarmAware 2025 reinforced its broader objective: to enable programmable, secure, and collaborative deployments across the IoT–Edge–Cloud continuum, supporting systems that are not only autonomous and scalable but also verifiable and trustworthy.

More Information