Adaptive policies under temporally correlated degradation

Design update policies that adapt to realized degradation under temporally correlated degradation costs, rather than relying only on age-dependent distributions and independence assumptions.

Background

The paper’s main model assumes independent degradation costs, but an additional experiment considers an AR(1) degradation process. Under temporal correlation, realized degradation contains predictive information about future degradation, so an optimal policy could use observed degradation histories rather than only the model age.

The authors report that the existing algorithms remain empirically effective under correlation, but they do not develop policies that exploit this predictive information. Designing such history- or context-adaptive policies remains unresolved.

References

We note that under correlation, the realized degradation predicts future degradation, so an optimal policy should adapt to the realized degradation. Designing such policies is left to future work.

— Learning When to Update: A Near-Optimal Timing Bandit Approach  (2609.37932 - Lin et al., 29 Sep 2026) in Section 6.2, Results and Discussion (commented robustness-to-temporal-correlation discussion)