A Formal translation from Robosim to C++ applied in the Robocup 2D simulation environment

Abstract

In the robotics field, simulation environments are essential for designing, evaluating, and validating complex autonomous behaviors before actual deployment, and the RoboCup 2D Simulation League provides a robust framework for testing multi-agent coordination and decision-making under uncertainty, benefiting AI and robotics research. Domain-specific languages like RoboSim enable intuitive high-level robot behavior modeling using finite-state machines (FSMs), yet converting these models into executable code is challenging due to the semantic gap between declarative DSL structures and imperative languages such as C++. This research presents an Eclipse plugin that automates the translation of RoboSim models into code using a C++ library that supports state machine implementations. The translation is systematic as it is based on defining and implementing translation rules to generate accurate FSM representations for the RoboCup 2D simulation environment. The approach ensures behavior consistency, facilitating the transition from high-level design to implementation and simplifying the development of autonomous agents while minimizing errors during deployment.