The AI@EDGE Architecture provides capabilities that go beyond the specific requirements of AI@EDGE use cases, aiming to support the following topics, which have been addressed with respective AI@EDGE architecture contributions:
- artificial intelligence “in-platform” inclusion (Network and Service automation intelligence) enabling better usage of infrastructure resources, AIFs as a concepts and interfaces, AIF descriptor;
- artificial intelligence “on-platform” inclusion (End-user Application intelligence); AIFs as a concepts and interfaces, AIF descriptor;
- data pipeline and data governance framework, being the data-driven platform, while preserving privacy and security of data in multi-stakeholder environment; AI/ML data pipeline including model lifecycle management and data collection components, aiming at an underlying data-driven architecture;
- E2E overall system orchestration and management, with orchestration and management for AIFs workflows. A NSAP architecture with new orchestration components, such as MTO, IOC and IARM; a CCP architecture with MEO, IOC and IARM; and new interfaces such as MEO to MEO and MTO to MEO.