Coupled Objects are an innovative way to ensure integrity of group of objects, or complex objects made of parts. This principle can be used in various applications such as logistic or security. The main property of coupled objects is that integrity checking is autonomous and does not depend on external information systems: all the necessary data are self contained in radio-frequency identification tags associated to the objects. This avoids important issues such as scalability and privacy, but the self-contained approach makes error diagnostics difficult when an integrity check fails. In this paper, we propose a solution to this problem, with a resilient data structure supporting the identification of missing elements in a coupled object. When some elements among the coupled objects are missing, it is possible to detect if the group is corrupted. Moreover, our approach also allows to identify the missing elements.