A domain model for automatic task planning in industrial logistics

• Grant number: 14/15831-8
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: October 01, 2014
• End date: September 30, 2015
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Marilza Antunes de Lemos
• Grantee: Felipe Guerra Soares
• Host Institution: Universidade Estadual Paulista (UNESP). Campus Experimental de Sorocaba. Sorocaba , SP, Brazil

Abstract

This project is inserted in the mobile robotics research area applied to industrial logistics, using concepts from automatic planning to develop a model of the idealized domain, which includes the manufacturing of industrial parts and that will be based on the Logistics League Sponsered Festo competition that occurs in Robocup events. The proposed model will be developed with the aid of itSIMPLE tool, an environment that offers a variety of features for creating models, which is not only possible modeling planning domains, but also testing and analysis of the model developed with several classic planners of the literature, maintenance of the created model, assisting in solving the problem and enabling intuitive and structured modeling process. The model should take into account the main requirements of the simulated domain of the competition, which is very similar to an industrial plant where a team composed of three Robotinos (Festo mobile robots) has to overcome the logistical challenges of a production system. The main challenge is to execute production plans that contain orders for different products. To produce, the robots must move in the arena toward the machines, in the sequence and correct time defined by product type. This requires a solution-plan that meets constraints set such as busy machine, machine under maintenance, availability of the robots, and time of sub-processes, among others. The basic task is the transport of materials, aimed not only at the fulfillment of the production plan established as challenge to the teams but also to minimize the time of the global task and the use of robots. The model will be built in ItSimple on an incremental approach, by implementing UML diagrams, creating agents that manipulate objects, as well as actions that can be taken by these entities, their attributes, and associations. The incremental approach provides tests of incomplete models with planners since the overall problem is complex. A sequence of problems will be defined to perform the tests on the complete model. For this, goal states and initial states will be specified. The problems will be submitted for each of the planners, and the plans generated, analyzed taking into account the metrics presented by itSIMPLE.