During the last few decades, maintenance for multi-component
systems have been extensively discussed in the literature. A
number of papers have been categorized and reported in [1–4].
These reviews show that most of works usually take advantages of
positive economic dependencies to minimize maintenance costs
by developing opportunistic maintenance policies [5–7] and
grouping maintenance policies. Positive dependencies refer to
situations in which the policies lead to maintenance setup costs
or/and downtime costs saving by performing simultaneously
several preventive maintenance activities. These reviews show
also that the maintenance grouping policies can be classified on
the basis of the planning horizon type: infinite horizons (stationary
grouping models) and finite horizons (dynamic grouping
models). The stationary models provide static rules for maintenance
which cannot be changed neither updated. They are mainly
based on intervention frequencies [8–10] or on control-limits
policies [11–15]. In the dynamic context, some dynamic grouping
models have been proposed to update and change the decision
rule according to available short-term information (varying