MONA LISA

Architecture

MONA LISA Architecture

MONA LISA connects data from models, software and hardware across the whole lifecycle of cyber-physical systems, and turns it into insight through model re-engineering, AI-supported analysis and visual analytics.

Overview

From use case data to insight: the MONA LISA processing chain.
From use case data to insight: the MONA LISA processing chain.

The industrial use cases provide three kinds of data: simulation data from models, logs and traces from software, and metered data from hardware. MONA LISA brings this data together and makes it usable across the lifecycle, from early design to operation and maintenance.

  • Unified trace storage. Data from models, hardware and software is stored in one place and converted to a common trace format, the Google Trace Event Format, so that all tools can work on it.
  • Model and domain knowledge re-engineering. Models of the system behaviour are reconstructed from trace logs with algorithmic and machine-learning methods.
  • AI-supported semantic analysis and integration. Large language models connect documents such as requirements, designs and test reports with the run-time data, supporting root cause analysis and predictive analytics.
  • Visual analytics. The results are presented to engineers, who can explore them from an overview down to the details.

Work package structure

Work package structure of MONA LISA.
Work package structure of MONA LISA.

The technical work follows the same chain. WP3 Infrastructure collects and converts the data, WP4 Model re-engineering builds models from it, WP5 Semantic analysis applies AI-based analysis, and WP6 Visual analytics presents the results to engineers.

WP2 Use cases, integration and demonstration brings the industrial use cases into the project and validates the results in real environments. WP7 Community building, dissemination and exploitation shares the results with industry and research, and WP1 Project management coordinates the consortium.