MONA LISA (Monitoring and Analytics for the whole Lifecycle, on Models, Hardware, and Software) is an ITEA 4 project that aims to connect engineering and operational data of cyber-physical systems across their entire lifecycle. This first newsletter summarises the activities carried out and the results achieved between January and June 2025.
Project start and consortium
Following the confirmation of the national funding decisions, the participating partners and countries were finalised. With the Turkish and Hungarian consortia confirmed alongside the Swedish partners, the project was formally launched with a consortium of academic and industrial partners from Sweden, Hungary and Türkiye.
The project kick-off meeting was held at KTH Royal Institute of Technology in Stockholm. During the meeting, the partners presented their roles and planned contributions, and the work plan, management structure and communication procedures were agreed. Since the kick-off, three plenary workshops and project meetings have been held in a hybrid format, complemented by bi-weekly teleconferences. The work plan was adjusted to the final composition of the consortium, and the preparation of the Project Consortium Agreement was started on the basis of the ITEA template.
Industrial use cases
The industrial use cases form the basis of all technical work in MONA LISA. They cover railway systems, steel production, laser research infrastructure, bridge monitoring and thermal imaging. During the reporting period, the use case providers started to refine their use cases and to define the stakeholder goals, requirements and success criteria that will guide the technical work packages. The evaluation of the use cases is led by Alstom.
Technical progress
Run-time tracing in Model RealTime and Code RealTime
HCL extended the run-time library of Model RealTime and Code RealTime so that applications developed with these tools can capture execution traces at run time. Tracing can be enabled when an application starts or while it is running. Two trace formats are supported: a native format with syntax colouring and navigation in Code RealTime, and a format for visualising traces as sequence diagrams on a web page.
Language models for engineering documentation
A master's thesis carried out at Alstom investigated the use of retrieval-augmented generation and large language models for searching and analysing internal technical documentation. Existing approaches were reviewed, and a prototype was built and evaluated on real documents.
New data types in Eclipse Trace Compass
KTH implemented the conversion of a new data type, smart contract transactions, into a format that the open-source Eclipse Trace Compass can read and visualise. The work provided the detailed understanding of the tool's architecture that is needed to add the data types required by the MONA LISA use cases.
Automated requirement quality checks
Mälardalen University integrated the NALABS requirement-smell detector into CI/CD pipelines. The tool now provides machine-readable output, quality thresholds and non-interactive execution, so that natural-language requirements are validated automatically at every commit.
Next steps
- Collection of shareable example data from the use case providers for the development of the data processing and visual analytics pipeline.
- Conversion of the partners' data to common trace formats for use with Eclipse Trace Compass.
- Selection of the models of computation and the tools to process them, reusing existing tools wherever possible.
- Definition of workflows to extract models from specification documents, taking different confidentiality levels into account.