Characterizing multi-contributor causal structures for planning
Proceedings of the first international conference on Artificial intelligence planning systems
Conditional nonlinear planning
Proceedings of the first international conference on Artificial intelligence planning systems
A critical look at Koblock's hierarchy mechanism
Proceedings of the first international conference on Artificial intelligence planning systems
A theory of conflict resolution in planning
Artificial Intelligence - Special volume on constraint-based reasoning
Acquiring search-control knowledge via static analysis
Artificial Intelligence
IJCAI'77 Proceedings of the 5th international joint conference on Artificial intelligence - Volume 2
Hi-index | 0.00 |
An important aspect of partial-order planning is the resolution of threats between actions and causal links in a plan. We present a technique for automatically deciding which threats should be resolved during planning, and which should be delayed until planning is otherwise complete. In particular we show that many potential threats can be provably delayed until the end; that is, if the planner can find a plan for the goal while ignoring these threats, there is a guarantee that the partial ordering in the resulting plan can be extended to eliminate the threats. Our technique involves: 1) construction of an operator graph that captures the interaction between operators relevant to a given goal, 2) decomposition of this graph into groups of related threats, and 3) postponement of threats with certain properties.