(2 min read)
During the first project year, Swarmchestrate has achieved significant progress in developing decentralized solutions for trust and knowledge management, particularly through the efforts of the Knowledge Management Workgroup, a group of experts from project partners, brought together to reshape how trust is established, maintained, and managed in a distributed digital ecosystem.
Key to these developments are three distinct approaches to trust reasoning:
- Accumulating trust, based on feedback and interaction history;
- Contextual trust, derived from the runtime environment; and
- Reputation-based trust, where entities’ performance and reliability are transparently evaluated.

Trust and identity management in Swarmchestrate (© Swarmchestrate consortium 2024-2026)
At the core of these advancements lies the global trust graph, a dynamic system where trust attributes like reliability, availability, reputation, and data quality are continuously updated based on real interactions. This mechanism empowers entities —be them clouds, swarms, artificial intelligence agents or resource providers— to autonomously manage their trust relationships, removing dependence on central authorities. Naturally, while some information is shared, each of the entities is allowed to form and use its own notions of trust. Albeit not planned for the upcoming implementation, an interesting idea for future extension is the integration of smart oracles; these may enhance this picture by performing complex trust calculations, interpreting policies, and ensuring that decisions are transparent and evidence-based. The use of reasoners that can operate directly on the global trust graph is still under consideration.
A highlight of the first year has been the development of ontology-driven knowledge management. Our initially crafted ontology provides a structured framework linking entities, trust relationships, and dynamic attributes, and aims at facilitating trusted interactions among the entities of the decentralized environment. This ontology, combined with a blockchain and a decentralized Knowledge Base (KB), ensures that metadata and interaction data are securely stored and accessible, supporting intelligent orchestration mechanisms at runtime.
The Knowledge Management Workgroup has also made strides in privacy and identity management, designing Self-Sovereign Identity (SSI) systems underpinned by Decentralized IDentifiers (DIDs), as an emerging W3C standard. These innovations not only secure user identities but also introduce Sybil-resistant mechanisms to prevent duplicate or fraudulent identities. Additionally, resource-specific sub-identifiers offer granular control and traceability, crucial for swarm-based orchestration scenarios.
The work on privacy-preserving computation has been equally interesting. By employing cutting-edge encryption methods like Functional Encryption (FE) and Hybrid Homomorphic Encryption (HHE), Swarmchestrate enables sensitive data to be analyzed securely within a distributed framework. This ensures confidentiality while facilitating artificial intelligence/machine learning and decision-making processes.
As the project moves on to its second year, the focus will be on refining these systems, scaling up implementations, and conducting real-world validations to solidify Swarmchestrate’s role in transforming decentralized information, knowledge and trust management, that follows the decades’ long line of evolution of grid, cloud and fog information systems.
Editors: Vlado Stankovski and Petar Kochovski, University of Ljubljana, Faculty of Computer and Information Science