From TR 103 781: There is: - Class s4envi:PhysicalObject, aligned with the DUL top-level ontology. - Class s4bldg:PhysicalObject, aligned with the DUL top-level ontology. Exact same definition as in SAREF4ENVI, but different axiomatization: s4envi:PhysicalObject ⊑ ∀ s4envi:isContainedIn . s4envi:PhysicalObject, while s4bldg:PhysicalObject ⊑ ∀ s4bldg:isContainedIn. (s4bldg:PhysicalObject ⊔ s4bldg:BuildingSpace) - SAREF4INMA reuses s4bldg:PhysicalObject - OP s4bldg:contains has a different definition than s4envi:contains. - OP s4bldg:isContainedIn has a different definition than s4envi:isContainedIn. SAREF4INMA reuses some classes from SAREF4BLDG such as s4bldg:Building, s4bldg:BuildingSpace, s4bldg:PhysicalObject, and specialize them as s4inma:Factory, s4inma:Area, s4inma:Site, s4inma:ProductionEquipment.This is a case for promoting some of the most generic SAREF4BLDG concepts to SAREF Core. Proposal: 1. promote saref:PhysicalObject to SAREF Core. 2. explicit that saref:Device is a subclass of saref:PhysicalObject. Open question: should saref:FeatureOfInterest be a subclass of saref:PhysicalObject ? I assume this would be too restrictive.