mc_rtc

mc_rtc

(opens in a new tab) (opens in a new tab) (opens in a new tab) (opens in a new tab)

NOTICE: jrl.cnrs.frhttps://jrl.cnrs.fr/mc_rtc (opens in a new tab) is currently down. Please use jrl-umi3218.github.iohttps://jrl-umi3218.github.io/mc_rtc (opens in a new tab) while we work on a fix.

Introduction

mc_rtc is an interface for simulation and robot control systems. These systems should provide the state of a given robot (joints' values, sensor readings...) and in return mc_rtc will provide the desired robot's state (command). This is done through the mc_control::MCGlobalController class. This class does not perform control by itself but rather delegates this task to the mc_control::MCController derived objects that it holds. Writing a controller within the mc_rtc framework is done by writing a class that inherits from the mc_control::MCController base class and implements the required functionnality. The details of this process can be found in the project documentation.

mc_rtc is a super-set of SpaceVecAlg/RBDyn/Tasks libraries which provides a friendlier/easier interface for the user.

This framework assumes some familiarity with the SpaceVecAlg (opens in a new tab), RBDyn (opens in a new tab) and Tasks (opens in a new tab) although the libraries in the mc_rtc package abstract most of these libraries inner-workings. Unfamiliar users may refer to the SpaceVecAlg and RBDyn tutorials (opens in a new tab) and the SpaceVecAlg and RBDyn presentations (opens in a new tab) authored by Joris Vaillant.

The project website (opens in a new tab) provides a somewhat complete introduction of the framework functionalities.

Feel free to open an issue (opens in a new tab) for questions, feature requests or problems you might encounter.

Available interfaces

The following interface between mc_rtc and 3rd-party framework are available:

• mc_rtc_ros (opens in a new tab) contains a number of tools related to the framework as well as a simple "ticker" that allows to run controllers in an open-loop fashion;
• mc_openrtm (opens in a new tab) embeds mc_rtc into a RTC component, this can be used to control a robot that is using the OpenRTM framework, an example with the virtual JVRC1 robot is also provided;
• mc_udp (opens in a new tab) provides a thin UDP-based client/server. The client is a regular mc_rtc interface that gets data through the UDP socket. The server can easily be implemented to interface with your robot. A server to interface with OpenRTM components is provided.
• mc_vrep (opens in a new tab) interface with the VREP simulation software;
• mc_naoqi (opens in a new tab) interface with the NAOqi framework (Softbank Robotics robots);
• mc_franka (opens in a new tab) interface with the libfranka library for Franka Emika robots;
• mc_mujoco (opens in a new tab) interface with the MuJoCo (opens in a new tab) physics simulator;

Installing

Please see mc_rtc installation guide (opens in a new tab) for details.

mc_rtc also uses the following great 3rd party libraries but does not require you to get or install them:

• Qhull 2020.2 (opens in a new tab) under the Qhull license (opens in a new tab)
• RapidJSON (opens in a new tab) under the MIT license (opens in a new tab)
• mpack (opens in a new tab) under the MIT license (opens in a new tab)