Image
Published 01/09/2025
Egi magazine - undefined
interTwin DTE Core and Infrastructure Modules
Image
Daniele Spiga
Cloud Solution Architect at INFN - WP5 Lead
Image
Isabel Campos
Researcher at CSIC - WP6 Lead

While the core modules enable the development and management of Digital Twins, the infrastructure modules enable their implementation.

The interTwin Digital Twin Engine (DTE) Core Modules enable the development and management of data-driven and compute-intensive applications by providing capabilities such as workflow composition, data fusion, AI workflow and method lifecycle management, real time acquisition and data analytics, validation, verification, and uncertainty tracing for model quality.

Core components include the Infrastructure Manager for deploying Digital Twins, OSCAR for triggering real-time jobs in Kubernetes, and SQAaaS for automated quality assurance in ML workflows. Ophidia supports multidimensional data analytics with provenance tracking, while yProv manages workflow traceability in coordination with SQAaaS, openEO, and Ophidia. openEO provides an API for Earth Observation workflows, now enhanced for real-time data via OSCAR. To support scalable ML, the itwinai Toolkit offers an open-source solution for running ML workloads on HPC and cloud systems, featuring unified logging, provenance tracking, and quality metrics.

The interTwin Digital Twin Engine (DTE) Infrastructure Modules have the role to implement all the capabilities needed by Digital Twins to effectively exploit computing resources.

In summary: 

  • Provide software solutions to enable resources provisioning on a wide range of compute providers to implement a digital continuum: interlink 
  • Support data access, data management and to support real time data processing,  in a federated environment: Teapot, ALISE, FTS3, RUCIO, Onedata S3,
  • Provide services and tools to enable the automated storage and compute resources orchestration enhancing the automatic and intelligent identification of the best provider either based on static and dynamic metrics: APEL Accounting; Kubernetes Usage Accounting Probe, AI Based Orchestrator,

Related magazine news