As an alternative approach to computing offline hybrid classical plans and then dealing with discrepancies/surprises online during plan execution by monitoring and replanning, we propose a parallel offline hybrid method, called HCPlan. This method suggests extending hybrid planning beyond classical planning, by inheriting advantages of conditional planning to deal with contingencies due to incomplete knowledge and partial observability and by integrating feasibility checks not only for executability of actuation actions but also for executability of sensing actions.

HCPlan relies on modeling deterministic effects of actuation actions and non-deterministic effects of sensing actions in the causality-based action language C+. Branches of a hybrid conditional plan are computed in parallel using a SAT solver, where continuous feasibility checks are performed as needed.  We provide theoretical guarantees on the soundness and completeness of HCPlan. 

We develop a comprehensive benchmark suite and introduce new evaluation metrics for hybrid conditional planning. We evaluate HCPlan with extensive experiments in terms of computational efficiency and plan quality, and compare HCPlan with other related conditional planners and approaches to deal with contingencies due to incomplete knowledge. We further demonstrate the applicability and usefulness of HCPlan in service robotics applications, through dynamic simulations and physical implementations.

We refer the reader to our journal paper for further information: https://doi.org/10.1177/0278364920963783 .