Across the project, WP1 ensured effective coordination, quality assurance, and continuous monitoring, establishing a management and quality framework early (via the Project Handbook) and sustaining technical alignment through regular consortium/WP meetings and structured internal communication procedures for the full project duration. In parallel, monitoring, quality control, and ethics/GDPR-aligned practices were maintained through risk management, deliverable review processes, and ethics procedures supporting compliant project execution. WP2 established the project’s architectural and specification backbone: defining and motivating the use cases, eliciting requirements and KPIs, and delivering the overall architecture and high-level design in D2.1 with detailed mappings into D2.2 (DApp/DAO specifications), D2.3 (data federation interfaces and OASEES Data Products aligned with GAIA-X), and D2.4 (accelerator integration spanning Quantum/GPU/SNN). These specifications provided the stable reference for implementation and integration across the remaining technical work packages. Building on that baseline, WP3 operationalized the platform through the OASEES SDK and lifecycle tooling, enabling repeatable edge onboarding and workload execution. The SDK supports CLI-driven onboarding (e.g. initializing lightweight Kubernetes clusters via K3s, joining worker nodes, and registering devices), and connects clusters to the Marketplace/DAO so they become “ready to accept workloads.” WP3 also tied deployment to marketplace flows by pulling container images from IPFS and performing capability-based placement (e.g. selecting workers based on accelerator requirements) while automating Kubernetes deployment and service exposure.
In WP4, OASEES advanced both edge intelligence and connectivity enablers. On the intelligence side, the project developed and evaluated cloud-to-edge distributed learning foundations (including federated and swarm-learning directions) as part of a broader AI-as-a-Service approach. On the networking side, WP4 delivered and validated a continuum networking framework that combines Istio service mesh, SD-WAN capabilities, and lightweight Kubernetes distributions to provide consistent communication, security and observability across edge and cloud resources, including support for low-power/IoT-friendly connectivity such as MQTT and 5G RedCap. Starting from M10, WP5 integrated and validated the full stack across six proof-of-concepts, demonstrating interoperable, edge-enabled, DAO-governed pipelines spanning acquisition → on-device inference → telemetry → on-chain governance → actuation. The final integrated validation confirmed operation across highly diverse edge devices and domains (health, EV/grid, drone swarms, structural monitoring, smart factory robotics, and IoT energy), with DAO-driven orchestration and end-to-end data-product flows validated in every PoC. This included quantified mission-level results in UC3 (e.g. ~35% UAV flight-time extension for a 5G RedCap/SNN configuration vs a standard 5G/ANN setup) and practical deployability in structural monitoring environments (automated installation, docker-compose-to-Kubernetes conversion, telemetry collection, reliable inter-device communication, and DAO-based decision-making with API-driven actuation). Finally, WP6 consolidated project impact through dissemination, exploitation, and standardization work.
