The present embodiments relate to performing an asset inventory for an industrial control system. By way of introduction, the present embodiments described below include systems and methods for asset discovery, asset characterization and semantic analysis. Incremental asset discovery in an industrial control system is provided using semantic analysis of the underlying production process using semantic models, template ontologies, and data analytics (e.g., based on metadata, network data, process data, etc.). Using the semantic analysis, different asset discovery methods are used to target individual assets of the industrial control systems to identify and baseline the assets. An accurate asset inventory is provided without risking disruption of the production process, and the asset inventory includes assets from highly segmented and isolated networks by adopting a combination of discovery methods, intelligent agents, and data analytics.
