Scholarlink

The growing heterogeneity of the compute continuum—from IoT and edge to cloud and HPC—makes complex data pipelines difficult to design and operate. SPICE is a unified platform for smart data pipelines that are semantically enriched, AI-assisted, secure, and explainable. It introduces Abstract Data Pipelines, a Unified Knowledge Layer, and a Perceptive Pipeline Composer to help users create reusable pipelines without deep platform expertise, while multi-objective optimisation and smart data/resource management dynamically map and adapt executions to heterogeneous infrastructures. A layered architecture virtualises access to diverse resources and integrates with XaaS marketplaces and European data spaces. Five use cases in manufacturing, energy, and medical imaging demonstrate how SPICE lowers entry barriers and enables more responsive, trustworthy, and efficient data-driven applications across the compute continuum.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

SPICE Project: Smart Data Pipelines for CognItive ComputE Continuum

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Martin Bobák,
  • Ladislav Hluchý,
  • Martin Šeleng,
  • Štefan Dlugolinský,
  • Thomas Fahringer,
  • Zahra Najafabadi Samani,
  • Juan Aznar Poveda,
  • Jacek Kitowski,
  • Bartosz Kryza,
  • Renata Slota,
  • Alvaro Vazquez Rodriguez,
  • Carlos Giraldo Rodriguez,
  • Fedor Lehocki,
  • Peter Kissa

摘要

The growing heterogeneity of the compute continuum—from IoT and edge to cloud and HPC—makes complex data pipelines difficult to design and operate. SPICE is a unified platform for smart data pipelines that are semantically enriched, AI-assisted, secure, and explainable. It introduces Abstract Data Pipelines, a Unified Knowledge Layer, and a Perceptive Pipeline Composer to help users create reusable pipelines without deep platform expertise, while multi-objective optimisation and smart data/resource management dynamically map and adapt executions to heterogeneous infrastructures. A layered architecture virtualises access to diverse resources and integrates with XaaS marketplaces and European data spaces. Five use cases in manufacturing, energy, and medical imaging demonstrate how SPICE lowers entry barriers and enables more responsive, trustworthy, and efficient data-driven applications across the compute continuum.