Three Approaches to Solve the Petrobras Challenge: Exploiting Planning Techniques for Solving Real-Life Logistics Problems
Abstract
The application of the state-of-the-art techniques into the real-world application is one of the most difficult challenges of the today's science in general and it is, also within the planning community, one of the strongest motivations driving the innovations in this field. Petrobras domain, being an abstraction of a real-life problem of resource-efficient transportation of goods from ports to the petroleum platforms is a good example of a difficult problem standing on the borderline between planning and scheduling that requires application of the state-of-the-art techniques of the modern artificial intelligent.
In this paper we provide three different ways of modeling and solving this domain: by utilizing classical planning, temporal planning, and finally, single-player games and Monte-Carlo Tree Search.