Technische Universität München Robotics and Embedded Systems





Project Objective

Compliant, musculoskeletal robotic systems offer several advantages, especially in situations where human and robot work in close proximity. A musculoskeletal design takes inspiration from the mechanics of the human body. It makes extensive use of viscous-elastic materials to emulate the muscles and tendons which enhance safety, dexterity and adaptivity in uncertain environments. It also allows reducing body weight and developmental cost, while at the same time increasing design flexibility.

Although there are several research platforms available that employ this design, current systems utilize custom-made, complex hardware and software, which inhibits their use beyond robotics research in academic settings. In fact, most of these systems are custom designed and built by one research group and, as a result, are seldom in use by people other than the initial developers.

The MYOROBOTICS project aims to improve the quality and reliability of the hardware involved and to make musculoskeletal robots readily available to researchers working in robotics and other domains (e.g. cognition, neuroscience), educators and the industry. The approach taken utilizes a modular design, involving components that can be easily interconnected in different ways to achieve required forms and functionality. These components will be mass-producible and reproducible (leveraging rapid prototyping techniques), improving cost-effectiveness and facilitating the transition to the market. A software toolchain will be made available that will allow the assembly of a virtual musculoskeletal robot, the definition of control algorithms and high-level behaviours, the optimization of the controller's performance and the simulation of its interaction with the environment. Three different control schemes will be developed that will target the individual muscles, the joints and the entire body, respectively.

All the aforementioned components, both software and hardware (the latter as CAD designs, board schematics and part lists) will be bundled in the MYOROBOTICS toolkit that will be made available as open-source to the community at large.



Myorobotics is a European Project; we closely work together with the following partners:



[1] Michael Jäntsch, Steffen Wittmeier, Konstantinos Dalamagkidis, Guido Herrmann, and Alois Knoll. Adaptive Neural Network Dynamic Surface Control: An Evaluation on the Musculoskeletal Robot Anthrob. In Proc. IEEE International Conference on Robotics and Automation ICRA, pages 4347-4352, 2015. [ .bib | .pdf ]
[2] Michael Jäntsch, Steffen Wittmeier, Konstantinos Dalamagkidis, Guido Herrmann, and Alois Knoll. Adaptive Neural Network Dynamic Surface Control for Musculoskeletal Robots. In Proc. IEEE Conference on Decision and Control CDC, pages 679-685, 2014. [ .bib | .pdf ]
[3] Michael Jäntsch, Steffen Wittmeier, Konstantinos Dalamagkidis, Alexander Panos, Fabian Volkart, and Alois Knoll. Anthrob - A Printed Anthropomimetic Robot. In Proc. IEEE-RAS International Conference on Humanoid Robots (Humanoids), pages 342-347, 2013. [ .bib | .pdf ]
[4] Michael Jäntsch, Steffen Wittmeier, Konstantinos Dalamagkidis, and Alois Knoll. Computed Muscle Control for an Anthropomimetic Elbow Joint. In Proc. IEEE/RSJ International Conference on Intelligent Robots and Systems IROS 2012, pages 2192-2197, 2012. [ .bib | .pdf ]