Artificial Intelligence for Safer Workplaces
CRISALIA investigates Artificial Intelligence methods and technologies for improving access to occupational health and safety knowledge. The project combines advanced document processing, Large Language Models, Retrieval-Augmented Generation and Knowledge Graphs to develop intelligent, reliable and traceable tools for accessing complex and continuously evolving regulatory and technical documentation.
About CRISALIA
AI technologies supporting occupational health and safety
Occupational health and safety relies on a large and heterogeneous body of regulations, technical documentation, guidelines and operational procedures that continuously evolves over time.
CRISALIA explores how Artificial Intelligence can make this knowledge easier to acquire, organise, analyse and query while preserving traceability to authoritative sources.
The project investigates integrated architectures combining document databases, semantic analysis, vector databases, Knowledge Graphs, Retrieval-Augmented Generation, intelligent agents and specialised conversational systems.
Particular attention is devoted to source verification, transparency, uncertainty management and the reliability of AI-generated answers.
Project vision
From documents to trusted knowledge
CRISALIA aims to transform fragmented occupational health and safety information into a continuously updated, structured and verifiable knowledge ecosystem supporting both human users and intelligent software agents.
Project indicators will be updated as research outputs become publicly available.
Hosted by
TeraLab · Sapienza University of Rome
CRISALIA is hosted by TeraLab, a research laboratory at Sapienza University of Rome working on High-Performance Computing, Artificial Intelligence and data-intensive applications.
Contact
Follow the project
Follow CRISALIA through TeraLab channels for project updates, research results, publications, software releases and events.