DeepRuler: Design, execute and evaluate path-following for robotic unmanned vehicles and experiment repeatability
Andrea Sorbara, Andrea Ranieri, Enrica Zereik, Marco Bibuli, Gabriele Bruzzone, Massimo Caccia · 2015
In robotics, nowadays, there is a strong evergrowing interest in executing replicable experiments to assess robots performance. This demand can't be easy fulfilled without established ad-hoc procedures to be followed. To this aim, an innovative software framework named DeepRuler is proposed. The presented software is able to design, automatically execute and evaluate highly replicable path-following experiments, minimizing the burden for the human operator of the robot. Among the many advantages of DeepRuler, there is its ease of integration with almost every kind of robot or simulator, without much effort left on the shoulders of the software developer. Another key feature is its open-source nature and its plugin-based structure, that allows the community to expand its functions in a very modular way.