An integrated model of planning and scheduling: main advantages and challenges

In this paper we properly analyze the main advantages and challenges of a novel model that integrates planning and scheduling. Both are artificial intelligence processes for tackling real-world problems. Planning studies action-selection for obtaining good quality plans. On the other hand, scheduling studies the suitable resources application in any plan. However, when the AI planning community has tried to solve real-world problems, they have seen the necessity to combine both planning and scheduling techniques. Additionally, in the state-of-the-art, there are two approaches to solve these problems: extended planning and extended scheduling. The first approach results in a high computational complexity, and the second approach uses a rigid plan template (embedded planner) avoiding the possibility of re-planning. Otherwise, our system homogeneously solves resources conflicts, in a flexible design with a strongly coupled form. Consequently, this model promises to be efficiently enough to obtain executable and consistent plans.