Data-driven Digital Evolution of Modern Laboratories

Abstract

The digital evolution of modern laboratories requires moving beyond isolated data capture toward machine-actionable representations of entire experimental workflows. This means formally modelling not only measurement data, but also process plans, execution details, and provenance. This talk presents a data-driven approach to laboratory digitalisation based on semantic technologies, in particular ontologies and knowledge graphs.

Ontologies provide a structured, interoperable representation of laboratory processes that captures both intended procedures and their actual execution. Such explicit provenance modelling is essential for reproducibility, systematic error analysis, outlier detection, and predictive modelling. By integrating process descriptions with resulting data and performance metrics, semantic models create a coherent knowledge layer that supports traceability, comparative analysis, and data-driven decision-making in diverse laboratory settings.

Addressing practical challenges of scalability, interoperability, and usability, knowledge graphs serve as flexible and scalable data infrastructures, while alignment with established domain ontologies ensures seamless integration across systems. A reusable and extensible reference model builds upon and integrates existing standards rather than replacing them, providing adaptable modelling patterns for diverse laboratory contexts. This framework establishes a pragmatic foundation for FAIR laboratory ecosystems and supports the transition toward data-driven research and innovation.