“…The performance of a diagnostic system is highly dependent on the set K and, consequently, dependent on the set K Σ , which highly depends on the dataflows, i.e., on the observations. Additional sensors lead to addtional constraints in K Σ and, therefore, to new sets in K. K can be obtained from combinations of constraints from K Σ using possible conflict generation (Pulido and Alonso, 2002), a bipartite graph (Blanke et al, 2006), the Dulmage-Mendelsohn decomposition (Krysander, Aslund and Nyberg, 2008) or elimination rules (Ploix, Désinde and Touaf, 2005). Basically, once K has been generated, it is possible to compute the performance of the diagnostic system in terms of detectability, discriminability or discernability, and diagnosability.…”