HOOPS Communicator supports basic collision detection, though it should not be utilized as a substitute for a full detailed analysis where absolute accuracy of this information is crucial. The collision detection algorithm uses the vertices from the tessellated model representation to compute minimum distances between bodies, faces, edges, etc. If the tessellation was generated from the B-Rep definition, your measurements will depend on the generated tessellation (i.e. a low LOD tessellation may result in a less accurate measurement since it is a less frequent sampling of the B-Rep). The algorithm will also simply return the minimum distance between two selected entities. It has no knowledge if one body is encapsulated by the other, etc. If you require a more detailed collision detection algorithm, please look into our HOOPS Exchange toolkit.