Towards Robust Execution of Mission Plans for Planetary Rovers

The paper discusses a novel methodology for robust plan execution involving an intelligent agent called Active Supervisor (ActS). ActS aims at preventing (at least in some cases) the failure of durative actions by anticipating anomalous trends of execution and by properly handling them. To reach this result, ActS needs that the plan to be executed carries some important pieces of knowledge: besides preconditions and effects, actions must also be described by intermediate conditions (i.e., invariant conditions), which must be satis- fied during the whole execution of durative actions. Tis knowledge is used by ActS to detect anomalous situations that may endanger the safeness of the plan executor. Whenever an anomaly has been detected, ActS tries to prevent a failure by changing the execution modality of the current action while it is still in progress. Preliminary experimental results, obtained in a simulated space exploration scenario, are reported.