Exploiting Search in Symbolic Numeric Planning with Patterns

The paper presents a new procedure for numeric planning using Symbolic Pattern Planning (SPP), which defines a formula encoding executable action subsequences from a starting state. The approach incrementally searches for intermediate states closer to a goal by dynamically recomputing and refining action patterns at each planning step, ensuring correctness and completeness under specific conditions. This method enhances the efficiency of numeric planning by leveraging symbolic search strategies, which is significant for practitioners aiming to optimize planning algorithms in AI applications.