The aim of this study is to investigate how semantic descriptions can facilitate interoperability, and to suggest an overall infrastructure for automatic information management in support of different logistic activities. Information models (ontologies) describe knowledge in a formal way, and can be used to enable interoperability between applications on a semantic level. OWL, Web Ontology Language, is a language in which such information models can be created. Semantic descriptions of web services enable automatic discovery, composition and execution. OWL-S, OWL-Services, is a vocabulary in which semantic descriptions of web services can be created. To enable automatic discovery a registry for web services that can perform searches based on semantic descriptions, and not only key words, is also needed. Complex information needs can be met through automatic composition, without having to pre-define processes of web services. Automatic discovery and composition are two very difficult problems, and different approaches to solve them have been investigated in this study. An infrastructure for automatic information gathering from heterogeneous data sources has also been suggested. The data sources are available as semantic web services in this. The core consists of a registry and a composer that can automatically discover and compose web services. A scenario with an international operation has been developed, and based on this a prototype of the infrastructure has been developed.