Model-based diagnosis (MBD) tackles the problem of troubleshooting sys
tems starting from a description of their structure and function (or b
ehavior). Time is a fundamental dimension in MBD: the behavior of most
systems is time-dependent in one way or another. Temporal MBD, howeve
r, is a difficult task and indeed many simplifying assumptions have be
en adopted in the various approaches in the literature. These assumpti
ons concern different aspects such as the type and granularity of the
temporal phenomena being modeled, the definition of diagnosis, the ont
ology for time being adopted. Unlike the atemporal case, moreover, the
re is no general ''theory'' of temporal MBD which can be used as a kno
wledge-level characterization of the problem. In this paper we present
a general characterization of temporal model-based diagnosis. We dist
inguish between different temporal phenomena that can be taken into ac
count in diagnosis and we introduce a modeling language which can capt
ure all such phenomena. Given a suitable logical semantics for such a
modeling language, we introduce a general characterization of the noti
ons of diagnostic problem and explanation, showing that in the tempora
l case these definitions involve different parameters. Different choic
es for the parameters lead to different approaches to temporal diagnos
is. we define a framework in which different dimensions for temporal m
odel-based diagnosis can be analyzed at the knowledge level, pointing
out which are the alternatives along each dimension and showing in whi
ch cases each one of these alternatives is adequate. In the final part
of the paper we show how various approaches in the literature can be
classified within our framework. In this way, we propose some guidelin
es to choose which approach best fits a given application problem. (C)
1998 Elsevier Science B.V. All rights reserved.