Portal de Periódicos da CAPES

Flexible Robot Strategy Design Using Belief-Desire-Intention Model

2011; Springer Science+Business Media; Linguagem: Inglês

10.1007/978-3-642-27272-1_5

1865-0937

Loris Fichera, Daniele Marletta, Vincenzo Nicosia, Corrado Santoro,

Auction Theory and Applications

This paper describes PROFETA, a Python framework developed by the authors to write robot strategies by means of the Belief- Desire-Intention (BDI) programming paradigm. This paradigm has been proposed in the field of autonomous agents programming and can be successfully applied also to autonomous robots thanks to their behavioural similarity with software agents. The paper describes the BDI model and AgentSpeak, a formal declarative language suitably designed for BDI agents. Then it introduces PROFETA, which takes inspiration from AgentSpeak and is designed with the objective of adding declarative constructs (needed by a BDI model) to an object-oriented and imperative language like Python. The result is a flexible environment that combines the power of both the classical object-oriented paradigm—useful for algorithm and control loop programming—and declarative approach—useful for AI and strategy programming. A case-study, based on Eurobot 2010 competition, shows such abilities, highlighting the main characteristics and advantages of PROFETA in strategy design.