Bulletin of the American Physical Society
APS March Meeting 2020
Volume 65, Number 1
Monday–Friday, March 2–6, 2020; Denver, Colorado
Session U22: Robophysics IIFocus Session
|
Hide Abstracts |
Sponsoring Units: DBIO Chair: Shai Revzen, University of Michigan Room: 303 |
Thursday, March 5, 2020 2:30PM - 2:42PM |
U22.00001: Centipede locomotion in varying terrains Alexandra Carruthers Ferrero, Kelimar Diaz, Yasemin Ozkan-Aydin, Joseph Mendelson, Daniel I Goldman Animals which move via waves of body undulation and/or few numbers of limbs have been extensively studied.Limited progress has been made studying animals with many degrees of freedom in their bodies in coordination with numerous limbs (i.e. myriapods like centipedes). To discover principles by which body undulation and limb dynamics in concert with body mechanics leads to high performance in diverse environments,we challenged Scolopocryptops sexspinosous (body length of ~4.5 cm, 23 pairs of legs) to traverse both a flat terrain and a Gaussian distributed rough terrain (12 x 12 cm with variable height of blocks). On flat terrain, centipedes maintain a constant out of phase traveling wave along their limbs during locomotion. In contrast, on rough terrain, centipedes will symmetrically use their limbs when their various segments are at different heights. We hypothesize that the centipedes will modulate the use of their limbs in order to overcome heterogeneities. Insight from a robophysical model suggests centipede limbs integrate passive dynamics allowing the locomotorto traverse myriad environments. |
Thursday, March 5, 2020 2:42PM - 2:54PM |
U22.00002: Modeling Locomotion in a Segmented Soft Robot using Planar Discrete Elastic Rods Nathaniel Goldberg, Xiaonan Huang, Carmel Majidi, Alyssa Novelia, Oliver M. O'Reilly, Derek A. Paley, William Scott Modeling soft robots that move on surfaces is challenging from a variety of perspectives. Recently Bergou, et al. introduced a numerically efficient formulation of Discrete Elastic Rods (DER) based on discrete differential geometry. In this talk, we describe a simplified planar version of Bergou et al.’s theory and show how it can be used to model soft robots that are composed of segments of soft material folded and bonded together. To show the efficacy of DER, the formulation is used to describe and analyze the dynamics of a prototype caterpillar-inspired soft robot. The robot consists of six segments of a shape memory alloy actuated soft material. By controlling the timing and sequencing of the actuators, locomotion can be achieved by exploiting stick-slip friction and the variation in the normal force exerted by the ground plane on each contacting actuator. Successful modeling of the robot also entailed the development and implementation of procedures to prescribe the parameters for components of the soft robot. After describing these procedures, we discuss the comparison of our calibrated model to the experimental behavior of the caterpillar-inspired soft robot. The extension of the DER modeling to other soft robot designs will be discussed. |
Thursday, March 5, 2020 2:54PM - 3:06PM |
U22.00003: Geometric mechanics reveals optimal coordination of limb and body waves during centipede locomotion Baxi Chong, Perrin E Schiebel, Yasemin Ozkan-Aydin, Malcolm Brown, Alexandra C Carruthers, Jennifer Rieser, Simon Sponberg, Daniel I Goldman Running epimorphic centipedes display a characteristic body undulation. Elucidating the role of body-leg coordination is difficult because of the many degrees of freedom in the flexible body and numerous limbs. We decomposed the centipede locomotion into body undulation and leg stepping waves, and used the mathematical framework of geometric mechanics to study body-leg coordination. We characterized the body-leg coordination by the phase offset between two waves. We recorded high-speed video and tracked the motion of a centipede (Scolopendra polymorpha, 5 trials) running on a treadmill. Our theory predicted that the speed (in body lengths per cycle, BLC) was maximized when the body undulation (1.6 spatial waves) had a phase of 2.50 rad leading that of the leg wave (2 spatial waves). The theoretical prediction of centipede speed was 0.40 BLC, 0.10 BLC and 0.25 BLC with best, worst and no body undulation respectively. Experiments revealed that the centipede moved at 0.41±0.05 BLC with the phase offset 2.51±0.71 rad, in good agreement with our theoretical prediction. Our preliminary results suggest that the geometric approach captures centipede locomotion dynamics without including inertial effects. |
Thursday, March 5, 2020 3:06PM - 3:18PM |
U22.00004: Geometric Gait Optimization with a five-link wheeled snake Brian Bittner, Shai Revzen Geometric mechanics offers a powerful set of tools for understanding locomotion. In past work we used data-driven oscillator theory to design a sample efficient method for modeling geometric systems such as the Purcell swimmer. We extended this method to include the modeling of highly damped systems inhabiting the perturbed Stokes dynamical regime. Here, we present a case study on a physical five-link wheeled snake robot. Under each link, a pair of wheels were co-aligned to decrease friction in the direction along the link. We seeded the geometric gait optimizer with a zero displacement gait. After 9 iterations of 80 cycles, we produced a motion that achieved 45% body length per cycle translation motion, while optimizing over 84 parameters. Running the system at 0.5 Hz, this took 24 minutes. We noticed the final gait changes combinations of wheels that are contacting and not contacting the ground, a feature not observed in the first gait, suggesting that the optimizer successfully exploited hybrid features of the dynamics. |
Thursday, March 5, 2020 3:18PM - 3:30PM |
U22.00005: Directional Compliance in Snake Robot Obstacle-Aided Locomotion Tianyu Wang, Julian Whitman, Matthew Travers, Howie Choset This work investigates the role of compliance in high-degree-of-freedom snake robots. Prior work used geometric methods to reduce the dimension of the robot’s state space to efficiently control the robot: instead of controlling all of the robot’s degrees-of-freedom, we need only control two for meaningful motions. Recently, we adapted this technique to handle unmodeled terrain irregularities. We defined a compliant controller on the reduced state space, which used torque information measured at the robot’s joints. This improved the robot’s performance, but still the robot arrived at stuck configurations. To address this problem, we introduce a biologically inspired strategy called directional compliance, which selectively allows some portions of the robot to comply to the environment, while others remain stiff to push off for forward progression. Unfortunately, with pure directional compliance method, the robot can still get stuck. Therefore, we develop an estimator to determine the state of the robot and a controller that switches between compliance and directional compliance modes to help the robot get unstuck. We experimentally find that our method enables the snake robot to locomote more consistently than the pure compliant controller in obstacle-rich environments. |
Thursday, March 5, 2020 3:30PM - 3:42PM |
U22.00006: Body compliance helps oscillating snake robots reduce roll instability to traverse large steps Qiyuan Fu, Chen Li Snakes traverse a diversity of complex terrain at ease, an ability that snake robots still aspire to achieve. Despite progress in arboreal (or similar) and granular environments, a major knowledge gap is how to maintain stability on large, smooth obstacles like large rocks or steps that lack “push points” to grip or brace against. Our previous study (Gart et al 2019 JEB) discovered that kingsnakes traverse large steps stably by combining lateral oscillation for propulsion with cantilevering for bridging height change. Here, to understand stability principles, we developed a snake robot as a physical model and tested two hypotheses: (1) roll stability diminishes as step becomes higher; and (2) body compliance helps maintain contact and reduce roll instability. With similar gait as the animal, the rigid body robot was able to traverse steps as high as 40% body length, but traversal probability quickly diminished with increasing step height. A locomotor transition analysis revealed that this was mainly due to an increase in roll instability. Adding body compliance improved contact, reduced roll instability, and increased traversal probability by up to 40% while maintaining speeds. Our study demonstrated that maintaining contact is important for stable locomotion in complex terrain. |
Thursday, March 5, 2020 3:42PM - 3:54PM |
U22.00007: Exploring the role of passive mechanics in limbless locomotors via a novel robophysical snake. Marine C Maisonneuve, Perrin E Schiebel, Daniel I Goldman Animals like snakes use traveling body bends to move in terrestrial terrain. Previously, we discovered that passive body buckling, facilitated by unilateral muscle activation, allowed negotiation of sparse obstacles without additional control input in the snake (C. occipitalis) [Schiebel et al., 2019]. To explore the implications of this scheme we developed a robophysical model with passive mechanical flexibility. Most snake robots precisely control the position of each joint with a single actuator. In contrast, the actuation in our robot is modeled after biological snakes, in which pairs of muscles, one on each side of the spine, create body bends by contracting unilaterally. The robot has 8 joints and 16 motors, each joint has two motor-driven pulleys on opposite sides of the body. The pulleys increase curvature by shortening a wire attached to the adjacent joint. Opposite an active motor, the pulley is completely unspooled; pairs of motors can resist forces which would lengthen active wires but not those pushing them shorter, allowing the body to passively buckle during certain obstacle collisions. Preliminary tests demonstrate that a travelling sinusoidal wave of joint angles allows locomotion using wheels (which provide directional friction). |
Thursday, March 5, 2020 3:54PM - 4:06PM |
U22.00008: Amplitude modulation enhances obstacle negotiation for sidewinders Abdul Kaba, Jennifer Rieser, Veronica M Paez, Henry Astley, Joseph Mendelson, Daniel I Goldman
|
Thursday, March 5, 2020 4:06PM - 4:18PM |
U22.00009: Springs and Wings: Elastic Energy Exchange in Insect Flight Jeffrey Gau, James Lynch, Nick Gravish, Simon Sponberg In many insects, wing movements are generated indirectly via deformations of an exoskeletal shell. Estimates of power expenditure suggest that elastic energy recovery between wingstrokes may reduce flight power requirements. We tested three questions: 1) Can the thoracic shell provide significant energy return? 2) Does a simple damped elastic model describe bulk mechanical behavior? and 3) Are different thorax regions specialized for elastic energy exchange? We measured deformation mechanics by recording the force required to sinusoidally deform the exoskeleton over a wide frequency range. Elastic energy storage in the exoskeleton is sufficient to minimize power requirements and a structural (frequency-independent) damping model, not a viscoelastic one, describes bulk mechanical properties. We next performed complementary experiments on a structurally damped hemispherical shell. In contrast to the hemispherical shell, mechanical coupling between exoskeleton regions improved spring performance and local properties depended on global strain patterns. We found regions specialized for energy recovery with low dissipation, highlighting the specificity of exoskeleton regions for flight energetics. Finally, we consider the implications of resonance mechanics on flapping wing flight. |
Thursday, March 5, 2020 4:18PM - 4:30PM |
U22.00010: Springs and Wings: Robophysical investigation of unsteady flapping wing dynamics James Lynch, Jeffrey Gau, Simon Sponberg, Nick Gravish Flying insects are thought to achieve energy-efficient flapping flight by storing and releasing elastic energy in their muscles, tendons, and thorax. However, the dynamics of even simple elastic elements coupled to nonlinear, unsteady aerodynamic forces may make controlling flapping motions via muscle force inputs challenging. In previous work, we examined the resonance properties of a dynamically-scaled robophysical system consisting of a rigid wing actuated by a motor in series with a spring. In this talk, we describe experiments to study the control implications of operating spring-wing transmissions on and off resonance. We consider the response of systems to changing control inputs and external perturbations. We conduct robophysical experiments and model simulations measuring the response of elastic flapping wing systems to realistic control inputs and external perturbations that simulate environmental interactions like wind gusts or collisions. We vary mechanical parameters of the system and track results across a non-dimensional parameter space, enabling comparisons across length scales. The results suggest that series-elastic flapping wings designed for maximum efficiency experience commensurate losses to control authority but may be more robust to external perturbations. |
Thursday, March 5, 2020 4:30PM - 4:42PM |
U22.00011: Springs and wings: robophysical and biological investigations inform the design, energetics, and control space for effective flapping wing flight James Lynch, Jeffrey Gau, Simon Sponberg, Nick Gravish Flying insects and robots possess a wide range of elastic structures that may enable energy saving in flapping wing flight. However, there is little direct experimental evidence of flight energy reduction through spring-like elements. Furthermore, the implications of a resonant flight system, such as the material and mechanical demands of the thorax, and the control and perturbation responses, are poorly understood. Here we present an overview of how elastic structures in flapping wing flight can improve flight energetics while at the same time altering control capabilities. We present these arguments through a non-dimensional representation of flapping-wing systems, supported by biological and robophysical experiments. This approach enables us to compare the energetic and control implications for flapping wing systems with elastic structures across the relevant spatial scales of flying insects and robots. |
Thursday, March 5, 2020 4:42PM - 4:54PM |
U22.00012: Biologically inspired vision based aerial robot perching Haijie Zhang, Jianguo Zhao Limited flight time has constrained the application of the aerial robot. One possible solution is to learn from biological flyers (i.e. birds, flies) to perch on the desired objects. During perching, the aerial robots can still do surveillance related tasks with the desired height and orientation instead of hovering. Research results have shown that visual information named time-to-contact is used for biological flyers perching guidance. And this information can be extracted from consecutive images. It is widely used for robot motion control for autonomous braking and landing since it can realize zero contact velocity. However, for aerial robot perching, a non-zero contact velocity is usually required to make the perching mechanism functionalize. In this research, we propose an optimized two-stage time-to-contact based guidance law. With a properly designed time-to-contact controller, we finally realize the fastest perching with desired contact velocity. The research presented in this paper can be readily applied to the control of the aerial robot for perching with visual feedback, and can inspire more alternative forms of time-to-contact based planning and control for robotic applications. |
Thursday, March 5, 2020 4:54PM - 5:30PM |
U22.00013: Bioinspired Embodied Control of Locomotion in Complex Environments Invited Speaker: Robert Full Robophysics is an indispensable approach to discovering general principles of robots operating in complex environments. Fortunately, physics principles governing non-equilibrium processes where motion emerges by internal actuation in synergy with robotic physical models meeting real world environments can generate quantifiable, testable hypotheses of complex biological systems not otherwise possible. In biology, robophysics has led to simple system-level models (templates) that collapse staggering complexity, while exploring how behavior emerges from more representative models (anchors). A next step is to examine the parallel and serial composition of templates and their dynamic reconfiguration. Robophysics provides the opportunity to determine how: biologically embodied mechanical computation works with neural computation; stability occurs in complex environments using soft-matter physics; performance of multiple tasks results without change in structure; multifunctional hierarchical and heterarchical control networks emerge; properties of fail-safe and fault tolerant performance evolve; and how learning can occur using a reduced set of control modules. In turn, robophysics approach to biological locomotion can inspire the design of novel robots and applied mathematics. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2025 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700