A lightweight robotic arm with pneumatic muscles for robot learning

Versatile motor skills for hitting and throwing motions can be observed in humans already in early ages. Future robots require high power-to-weight ratios as well as inherent long operational lifetimes without breakage in order to achieve similar perfection. Robustness due to passive compliance and high-speed catapult-like motions as possible with fast energy release are further beneficial characteristics. Such properties can be realized with antagonistic muscle-based designs. Additionally, control algorithms need to exploit the full potential of the robot. Learning control is a promising direction due to its the potential to capture uncertainty and control of complex systems. The aim of this paper is to build a robotic arm that is capable of generating high accelerations and sophisticated trajectories as well as enable exploration at such speeds for robot learning approaches. Hence, we have designed a light-weight robot arm with moving masses below 700 g with powerful antagonistic compliant actuation with pneumatic artificial muscles. Rather than recreating human anatomy, our system is designed to be easy to control in order to facilitate future learning of fast trajectory tracking control. The resulting robot is precise at low speeds using a simple PID controller while reaching high velocities of up to 12 m/s in task space and 1500 deg/s in joint space. This arm will enable new applications in fast changing and uncertain task like robot table tennis while being a sophisticated and reproducible test-bed for robot skill learning methods. Construction details are available.

[1]  Mustapha Hamerlain,et al.  Experimental Study of Nonsingular Terminal Sliding Mode Controller for Robot Arm Actuated by Pneumatic Artificial Muscles , 2014 .

[2]  Koh Hosoda,et al.  Direct teaching method for musculoskeletal robots driven by pneumatic artificial muscles , 2012, 2012 IEEE International Conference on Robotics and Automation.

[3]  Bertrand Tondu,et al.  Modelling of the McKibben artificial muscle: A review , 2012 .

[4]  George Nikolakopoulos,et al.  Piecewise Affine Modeling and Constrained Optimal Control for a Pneumatic Artificial Muscle , 2014, IEEE Transactions on Industrial Electronics.

[5]  Blake Hannaford,et al.  McKibben artificial muscles: pneumatic actuators with biomechanical intelligence , 1999, 1999 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (Cat. No.99TH8399).

[6]  Shane Legg,et al.  Human-level control through deep reinforcement learning , 2015, Nature.

[7]  B. Tondu.  Robust and Accurate Closed-Loop Control of McKibben Artificial Muscle Contraction with a Linear Single Integral Action , 2014 .

[8]  Pierre Lopez,et al.  Modeling and control of McKibben artificial muscle robot actuators , 2000 .

[9]  Yanhe Zhu,et al.  One Nonlinear PID Control to Improve the Control Performance of a Manipulator Actuated by a Pneumatic Muscle Actuator , 2014 .

[10]  Bram Vanderborght,et al.  Proxy-based Sliding Mode Control of a Planar Pneumatic Manipulator , 2009, Int. J. Robotics Res..

[11]  Darwin G. Caldwell,et al.  Control of pneumatic muscle actuators , 1995 .

[12]  Kyoung Kwan Ahn,et al.  Design and implementation of an adaptive recurrent neural networks (ARNN) controller of the pneumatic artificial muscle (PAM) manipulator , 2009 .

[13]  Bertrand Tondu,et al.  A Seven-degrees-of-freedom Robot-arm Driven by Pneumatic Artificial Muscles for Humanoid Robots , 2005, Int. J. Robotics Res..

[14]  Nikolaos G. Tsagarakis,et al.  Pneumatic muscle actuator technology: a light weight power system for a humanoid robot , 1998, Proceedings. 1998 IEEE International Conference on Robotics and Automation (Cat. No.98CH36146).

[15]  A. Haith,et al.  Model-based and model-free mechanisms of human motor learning. , 2013, Advances in experimental medicine and biology.

[16]  Jan Peters,et al.  A Survey on Policy Search for Robotics , 2013, Found. Trends Robotics.

[17]  Oliver Kroemer,et al.  Learning to select and generalize striking movements in robot table tennis , 2012, AAAI Fall Symposium: Robots Learning Interactively from Human Teachers.

[18]  Andreas Schulz,et al.  A Human-Like Robot Hand and Arm with Fluidic Muscles: Biologically Inspired Construction and Functionality , 2003, Embodied Artificial Intelligence.

[19]  Kyoung Kwan Ahn,et al.  Nonlinear PID control to improve the control performance of 2 axes pneumatic artificial muscle manipulator using neural network , 2006 .

[20]  Oussama Khatib,et al.  Design and Control of a Bio-inspired Human-friendly Robot , 2010 .

[21]  Ching-Ping Chou,et al.  Static and dynamic characteristics of McKibben pneumatic artificial muscles , 1994, Proceedings of the 1994 IEEE International Conference on Robotics and Automation.

[22]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[23]  O. Sawodny,et al.  Cascaded control concept of a robot with two degrees of freedom driven by four artificial pneumatic muscle actuators , 2005, Proceedings of the 2005, American Control Conference, 2005..

[24]  Sang Wan Lee,et al.  The structure of reinforcement-learning mechanisms in the human brain , 2015, Current Opinion in Behavioral Sciences.

[25]  Blake Hannaford,et al.  Measurement and modeling of McKibben pneumatic artificial muscles , 1996, IEEE Trans. Robotics Autom..

[26]  B. Tondu,et al.  McKibben artificial muscle can be in accordance with the Hill skeletal muscle model , 2006, The First IEEE/RAS-EMBS International Conference on Biomedical Robotics and Biomechatronics, 2006. BioRob 2006..

[27]  Jan Peters,et al.  A biomimetic approach to robot table tennis , 2010, 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[28]  Minoru Asada,et al.  Real-Time Inverse Dynamics Learning for Musculoskeletal Robots based on Echo State Gaussian Process Regression , 2012, Robotics: Science and Systems.