UT Austin Villa RoboCup Humanoid League 2018

Team Leader: Justin Hart

Team Members: Justin Hart, Luis Sentis, Peter Stone

Upcoming Robotic System: Liquid Cooled Viscoelastic Robot Apptronik Draco. 10 Degrees of Freedom. Expected completion date is April 2018. Manufacturer: Apptronik Systems. Height (including head): 170 cm.

Disclaimer: Only the lower part of the humanoid is being shown here. Arms and a headsensor with a Multisense SL camera system will be incorporated at a later date before the competition. Batteries and onboard computer will also be incorporated before competition.

CAD Design of Draco

CAD Detail of Liquid Cooled Viscoelastic Actuators

Detail of performance liquid cooled actuators

Description of the approaches and information on scientific achievements:

• We will leverage UT Austin Villa’s advancement in the NAO SPL League including: Grounded Action Transformation for Robot Learning in Simulation, Ball Detection, Deep Reinforcement Learning in Parameterized Action Space, Layered Learning Strategies Applied to Individual Behaviors in Robot Soccer, Adaptation of Surrogate Tasks for Bipedal Walk Optimization, Long-Distance Kicking, Robot-centric Activity Recognition, Collision-avoiding Role Assignment, Humanoid Robots Learning to Walk Faster, etc.
• We will leverage the Human Centered Robotics Lab advancements on Closed-Loop Whole Body Operational Space Control, Robust Dynamic Locomotion via Reinforcement Learning, Building Biped Robots with Viscoelastic Liquid Cooled Actuators, Robust Optimal Planning and Control of Non-Periodic Bipedal Locomotion with A Centroidal Momentum Model, Interactive Whole-Body Inverse Kinematics, etc.

Relevant publications:

1. D.H. Kim, J. Lee, L. Sentis, Robust Dynamic Locomotion via Reinforcement Learning and Novel Whole Body Controller, Under Review for 2018 Publication
2. Y. Zhao, B. Fernandez, L. Sentis, Robust Optimal Planning and Control of Non-Periodic Bipedal Locomotion with A Centroidal Momentum Model, International Journal of Robotics Research, (public arXiv preprint), 36(11): 1211-1242, September 2017
3. D.H. Kim, S.J. Jorgensen, P. Stone, L.Sentis, Dynamic Behaviors on the NAO Robot With Closed-Loop Whole Body Operational Space Control, IEEE International Conference on Humanoid Robots (Humanoids 2016), November 2016
4. Josiah Hanna and Peter Stone. Grounded Action Transformation for Robot Learning in Simulation. In Proceedings of the 31st AAAI Conference on Artificial Intelligence (AAAI), February 2017.
5. Jacob Menashe, Josh Kelle, Katie Genter, Josiah Hanna, Elad Liebman, Sanmit Narvekar, Ruohan Zhang, and Peter Stone. Fast and Precise Black and White Ball Detection for RoboCup Soccer. In RoboCup-2017: Robot Soccer World Cup XXI, July 2017.
6. Matthew Hausknecht and Peter Stone. Deep Reinforcement Learning in Parameterized Action Space. In Proceedings of the International Conference on Learning Representations (ICLR), May 2016.
7. David L. Leottau, Javier Ruiz-del-Solar, Patrick MacAlpine, and Peter Stone. A Study of Layered Learning Strategies Applied to Individual Behaviors in Robot Soccer. In Luis Almeida, Jianmin Ji, Gerald Steinbauer, and Sean Luke, editors, RoboCup-2015: Robot Soccer World Cup XIX, Lecture Notes in Artificial Intelligence, Springer Verlag, Berlin, 2016.
8. Patrick MacAlpine, Elad Liebman, and Peter Stone. Adaptation of Surrogate Tasks for Bipedal Walk Optimization. In GECCO Surrogate-Assisted Evolutionary Optimisation (SAEOpt) Workshop, July 2016.
9. Mike Depinet, Patrick MacAlpine, and Peter Stone. Keyframe Sampling, Optimization, and Behavior Integration: Towards Long-Distance Kicking in the RoboCup 3D Simulation League. In Reinaldo A. C. Bianchi, H. Levent Akin, Subramanian Ramamoorthy, and Komei Sugiura, editors, RoboCup-2014: Robot Soccer World Cup XVIII, Lecture Notes in Artificial Intelligence, Springer Verlag, Berlin, 2015.
10. Ilaria Gori, Jivko Sinapov, Priyanka Khante, Peter Stone, and J.K. Aggarwal. Robot-centric Activity Recognition ‘in the Wild’. In Proceedings of the International Conference on Social Robotics (ICSR), October 2015.
11. Patrick MacAlpine, Eric Price, and Peter Stone. SCRAM: Scalable Collision-avoiding Role Assignment with Minimal-makespan for Formational Positioning. In Proceedings of the Twenty-Ninth AAAI Conference on Artificial Intelligence (AAAI), January 2015.
12. Alon Farchy, Samuel Barrett, Patrick MacAlpine, and Peter Stone. Humanoid Robots Learning to Walk Faster: From the Real World to Simulation and Back. In Proc. of 12th Int. Conf. on Autonomous Agents and Multiagent Systems (AAMAS), May 2013.

Relevant videos that show the skills of the team