Conditional temporal planning: how chance encounters and casual collaborations may raise great expectations
摘要
In 2003, we introduced the Conditional Temporal Problem (CTP) formalism as a novel approach to address the challenges of temporal constraints in conditional planning. This commentary revisits the origins of our collaboration, the key innovations of CTP, and its impact on subsequent research in temporal reasoning and planning. We discuss how CTP extended standard temporal constraint-satisfaction models by incorporating observation nodes and conditional labels, enabling the construction of plans that adapt dynamically to observations during execution. The paper also explores the past and current evolutions of the field, including extensions and open challenges, particularly in integrating symbolic AI with learning-based approaches and multi-agent systems.