Walking, Rolling, and Beyond: First-Principles and RL Locomotion on a TARS-Inspired Robot
Robotic locomotion research often draws inspiration from biologically inspired leg designs. However, many human-engineered settings can benefit from non-anthropomorphic forms. This paper presents TARS3D, a research platform that translates the block-shaped 'TARS' robot from the film Interstellar into a 0.25 m, 0.99 kg robot with seven actuated degrees of freedom. The film shows two primary gaits: a bipedal-like walk and a high-speed rolling mode. We build reduced-order models for each gait, derive closed-form limit-cycle conditions, and validate the predictions on hardware.
Experiments confirm that the robot respects its ±150° hip limits, alternates left-right contacts without interference, and maintains an eight-step hybrid limit cycle in rolling mode. The rolling gait is modeled as an eight-spoke double rimless wheel, with each telescopic leg providing four contact corners. The robot's telescopic leg redundancy implies a far richer gait repertoire than the two limit cycles treated analytically.
To explore this space, we use deep reinforcement learning (DRL) in simulation. The learned policy can recover the analytic gaits under the right priors and discover novel behaviors. Our findings show that TARS3D's fiction-inspired bio-transcending morphology can realize multiple previously unexplored locomotion modes. The combination of analytic synthesis and reinforcement learning opens a promising pathway for multimodal robotics.
References
[1] S. H. Collins, A. Ruina, R. Tedrake, and M. Wisse, "Efficient bipedal robots based on passive-dynamic walkers," Science, vol. 307, no. 5712, pp. 1082–1085, 2005.
[2] M. Raibert et al., "Bigdog, the rough-terrain quadruped robot," International Journal of Robotics Research, vol. 27, no. 3–4, pp. 339–353, 2008.
[3] U. Saranli, M. Buehler, and D. E. Koditschek, "Rhex: A simple and highly mobile hexapod robot," International Journal of Robotics Research, vol. 20, no. 7, pp. 616–631, 2001.
[4] M. Hutter et al., "Anymal—a highly mobile and dynamic quadrupedal robot," IEEE Robotics and Automation Letters, vol. 2, no. 4, pp. 2201–2208, 2017.
[5] A. Kumar and G. Dudek, "Terrain-adaptive reconfigurable robots," International Journal of Robotics Research, vol. 40, no. 8–9, pp. 927–943, 2021.
[6] R. Udell, "Breaking free from anthropomorphism: Purpose-first robot design," Robotics Today, 2022.
[7] J. Kim, "Bio-transcending robots: Beyond animal morphology," Ph.D. dissertation, University of Illinois at Urbana–Champaign, 2023.
[8] M. Bualat et al., "Astrobee: A free-flying robot for the international space station," in Proc. IEEE Aerospace Conf., 2019.
[9] A. Chakravarthy and K. Sreenath, "Ringbot: A leg–wheel hybrid robot for dense urban mobility," IEEE Robotics and Automation Letters, 2024.
[10] C. D. Nolan, N. P. D. Crowley, and S. P. E. S. Fisher, "Interstellar [motion picture]," Warner Bros. Pictures, Legendary Pictures, Syncopy, 2014, TARS robot character 2014 Warner Bros. Entertainment Inc. All rights reserved.
[11] J. Schulman, F. Wolski, P. Dhariwal, A. Radford, and O. Klimov, "Proximal policy optimization algorithms," 2017.
[12] M. Mittal, C. Yu, Q. Yu, J. Liu, N. Rudin, D. Hoeller, J. L. Yuan, R. Singh, Y. Guo, H. Mazhar, A. Mandlekar, B. Babich, G. State, M. Hutter, and A. Garg, "Orbit: A unified simulation framework for interactive robot learning environments," IEEE Robotics and Automation Letters, vol. 8, no. 6, pp. 3740–3747, 2023.
[13] H. Lipson and J. B. Pollack, "Automatic design and manufacture of robotic lifeforms," Nature, vol. 406, pp. 974–978, 2000.
[14] N. Cheney, J. Clune, H. Lipson, and J. C. Bongard, "Unshackling evolution: Evolving soft robots with multiple materials," Proceedings of the National Academy of Sciences, vol. 111, pp. 3228–3233, 2014.
[15] A. Cully, J. Clune, D. Tarapore, and J.-B. Mouret, "Robots that can adapt like animals," Nature, vol. 521, pp. 503–507, 2015.
[16] J. C. Bongard, V. Zykov, and H. Lipson, "Resilient machines through continuous self-modeling," Science, vol. 314, pp. 1118–1121, 2006.
[17] M. Garcia, A. Ruina, M. Coleman, and A. Chou, "The simplest walking model: Stability, complexity, and scaling," Journal of Biomechanical Engineering, vol. 120, no. 2, pp. 281–288, 1998.
[18] A. Goswami, B. Thuilot, and B. Espiau, "Compass-like biped robot: Symmetry and chaos in passive gait," International Journal of Robotics Research, vol. 17, no. 12, pp. 1282–1301, 1998.
[19] E. R. Westervelt, J. W. Grizzle, C. Chevallereau, and B. Morris, "Feedback controllers for rapid stabilization of periodic orbits in bipedal locomotion," IEEE Transactions on Automatic Control, vol. 48, no. 12, pp. 2153–2167, 2003.
[20] F. Asano and Z.-W. Luo, "Biped gait generation and control based on a 3-d extender passive dynamic walking model," IEEE Transactions on Robotics, vol. 24, no. 5, pp. 1234–1246, 2008.
[21] A. D. Kuo, "Energetics of actively powered locomotion using the simplest walking model," Journal of Biomechanical Engineering, vol. 124, no. 1, pp. 113–120, 2002.
[22] T. McGeer, "Passive dynamic walking," International Journal of Robotics Research, vol. 9, no. 2, pp. 62–82, 1990.
[23] K. Byl and D. E. Koditschek, "Mule: Stable, compliant rolling rimless-wheel robots," in Proc. 11th Int. Conf. on Climbing and Walking Robots, 2008, pp. 17–24.
[24] P. A. Bhounsule, M. Zambelli, A. Ruina et al., "Low-cost, record-setting robotic walking with Ranger," Science Robotics, vol. 9, 2014, article #e2014-09.
[25] J. Hwangbo, J. Lee, A. Dosovitskiy, C. D. Bellicoso, V. Tsounis, V. Koltun, and M. Hutter, "Learning agile and dynamic motor skills for legged robots," Science Robotics, vol. 4, no. 26, p. eaau5872, 2019.
[26] Z. Li, X. B. Peng, P. Abbeel, S. Levine, G. Berseth, and K. Sreenath, "Reinforcement learning for versatile, dynamic, and robust bipedal locomotion control," 2024.
[27] A. Sripada and H. Geyer, "Turning behavior of running systems induced by leg placement," Master’s Thesis, Carnegie Mellon University, The Robotics Institute, 2021.