A standardized formalism used to describe planning problems for automated planners. It provides a structured way to represent the initial state of the world, the desired goal state, and the possible actions that can be taken to transition between states. For instance, it can define actions like “pick up object,” specifying preconditions (object must be reachable) and effects (object is now held). This enables a planner to reason about sequences of actions that achieve a desired outcome.
This formalism is crucial for developing and evaluating artificial intelligence planning systems. Its adoption facilitates communication and comparison of different planning algorithms. By using a common language, researchers can easily share planning domains and test the performance of their algorithms on established benchmarks. Its development represents a significant step towards creating more general and reusable planning systems, moving beyond ad-hoc problem-specific solutions.
