In the case of ETTO, no unpredictable events are identified, we assumed that all the interactions where known in advance. Then there is no cooperation failure at the level of the interactions between entities and the system, but a possible cooperation failure can appear if the result presented by the system does not satisfy a constraint from an entity.

If we take another example of an application with physical sensors. These sensors can be prone to failures and thus they will deliver false data to the system. In fact, these actors are not reliable and it must be mentioned within the concerned use case.