Bulletin of the American Physical Society
APS March Meeting 2023
Volume 68, Number 3
Las Vegas, Nevada (March 5-10)
Virtual (March 20-22); Time Zone: Pacific Time
Session S10: Robophysics IIFocus
|
Hide Abstracts |
Sponsoring Units: DBIO Chair: Perrin Schiebel, Harvard University Room: Room 202 |
Thursday, March 9, 2023 8:00AM - 8:36AM |
S10.00001: Active and driven wave-propelled interfacial particles Invited Speaker: Daniel M Harris When an asymmetric floating body is internally or externally vibrated, the self-generated capillary wavefield can lead to steady propulsion or rotation. In this talk, I will discuss several related and recently discovered systems that leverage this physical mechanism. First, I will present our work on the "SurferBot": a centimeter-scale robotic device that self-propels along a fluid interface using an onboard vibration motor. I will then transition to a vibrating fluid substrate, where freely floating particles are shown to self-propel along straight paths, rotate in place, or move along curvilinear trajectories, depending sensitively on the particle asymmetries and driving parameters. Such particles interact at a distance through their mutual wavefield and exhibit a rich array of multibody dynamics. Overall, this highly accessible and tunable macroscopic system serves as a novel platform for exploring active and driven matter interacting in fluid environments. |
Thursday, March 9, 2023 8:36AM - 8:48AM |
S10.00002: The role of mechanical properties of the basilisk lizard foot in rejecting perturbations while running at the air-water interface Henry Cerbone, Michelle C Yuen, Perrin E Schiebel During escape, basilisk lizards, B. basiliscus, run rapidly (~7-21 body lengths/second) on the surface of water. Previous work elucidated how the trajectory of the lizard's foot through water generates propulsion [Hsieh 2004]. However, the role of the lizard's long, flexible toes has not been as closely studied. Body weight support primarily occurs from the stroke of the foot through the water with high Reynold's number (3.6 × 107) [Gaudet 1998]. We hypothesized that the properties of the foot, particularly the shape and stiffness, can passively correct for kinematic errors which would otherwise result in atypical perturbations to the reaction force. To test this, we developed several robophysical models of the foot varying in fidelity from simple 2D shapes cut from acrylic to 3D-printed and silicone cast feet based on biological scanning and testing of cadaveric specimens. These foot models were mounted to a robotic arm coupled with a load cell to measure the reaction forces during the interaction with a fluid medium. We began with a foot trajectory based on existing biological data before introducing error/noise into the path. We performed a sensitivity analysis of the reaction forces as a function of perturbation and foot model type, coupled with cavity-drag analysis to better characterize the stroke effectiveness of the foot models. This study allows us to begin to understand the role of mechanical properties in the foot in rejecting perturbations due to kinematic error. |
Thursday, March 9, 2023 8:48AM - 9:00AM |
S10.00003: Robot Doggy Paddle: Modeling and optimization reveal efficient swimming gaits Sean Gart, Max Austin, Jonathan Clark, Jason Pusey Deep water is common in many typical operating environments of legged robots. Despite this, study of legged robot gaits has mainly focused on terrestrial movement. Efficient swimming gaits are needed to greatly expand the operational capability of legged robots. Building on previous work on single leg swimming gaits, this study describes a 7 degree-of-freedom reduced-order dynamic model that captures surface swimming of a bipedal “sagittal quadruped” robot made of two rigidly connected “hip” masses with revolute-revolute legs connected at each hip. With the model, we investigated the effect of the gait parameters of stroke length and angle, leg extension and retraction, and front-rear phasing on the speed, efficiency, and stability of swimming gaits. To study the effect of body length, center-of-mass location, and buoyancy, on speed, efficiency, and stability of swimming, we used direct colocation optimization to find an optimal gait for each set of parameters. These results could be used to inform gaits and design parameters for existing and novel legged robots to extend their operational capabilities beyond terrestrial locomotion. |
Thursday, March 9, 2023 9:00AM - 9:12AM |
S10.00004: Magnetic, modular, undulatory robots as robophysical models for exploration of fish-inspired swimming. Hankun Deng, Donghao Li, Kundan Panta, Bo Cheng Evolution has successfully explored various forms of fluid-structure interaction (FSI) for underwater locomotion, from which a great diversity of fish swimming has emerged. However, it remains unclear how fish or fish-inspired swimming emerges from the complex physics of FSI which includes interactions among active structures (e.g., fins and elongated bodies), neuromuscular and motor control, and the physics of fluids. We address this problem using robophysical models of undulatory swimming (magnetic, modular, undulatory robots). Via experimental motor learning, we systematically explored the relationship among the body morphology, motor control, swimming gaits and performance. First, we studied the effects of the robot design (e.g., body length, caudal fin stiffness) on the emergence of swimming gaits and performance (e.g., forward/backward swimming, turning maneuver). Second, we identified the embodied properties in the FSI of the emergent gaits using motor learning, frequency response and motor control perturbation. Third, we employed time-resolved Particle Image Velocimetry (PIV) to quantitatively study the hydrodynamic properties for a diversity of emergent swimming gaits. These results provide novel insights and guiding principles for fish and fish-inspired robot swimming. |
Thursday, March 9, 2023 9:12AM - 9:24AM |
S10.00005: Modeling and Characterization of Bi-flagellated Robot with Tumbling Zhuonan Hao, Sangmin Lim, Mohammad Khalid Jawed Multi-flagellated bacteria locomote due to the interaction between two or more filamentary tails in low Reynolds number flow, i.e., bundling and tumbling. One of the fundamental goals of robots inspired by bacterial locomotion is to develop an effective and steerable robot based on the interplay of filamentary structure and hydrodynamics. In this work, we build a macroscopic bio-inspired robot with two rigid flagella (right-handed helices) connected to a cylindrical head. The robot body can reorient with repeatable and controllable tumbling when two flagella rotate in the opposite direction. We model this bi-flagellated mechanism interacting with the low Reynold fluid by coupling rigid body dynamics and the method of Regularized Stokeslet Segments. We further explore the parameter space of the simulation and the experiment to maximize the rate of change of reorientation. Moreover, we propose a run-and-tumble control scheme by modulating the rotation speed and direction of two flagella so that the locomotion gait is analytically tractable. We expect our framework to encourage more study and application on the mobility and controllability of micro-scale bacterial robots. |
Thursday, March 9, 2023 9:24AM - 9:36AM |
S10.00006: Data-driven discovery of hierarchical neuromechanical systems Benjamin McInroe, Yuliy Baryshnikov, Daniel Koditschek, Robert J Full Dynamic locomotion behaviors result from high dimensional, nonlinear, and dynamically coupled interactions between an animal or robot and its environment. The templates and anchors hypothesis resolves this complexity by postulating that preferred postural degrees of freedom for a specific task are anchored - rendered stable and robust to perturbations - by fast neural and mechanical feedback control forces. However, due to the lack of a general methodology for identifying templates from behavioral data, the use of templates to study dynamic locomotion behaviors in animals, and translate such behaviors to robophysical systems, has largely been limited to a few well-studied examples. Further, reliance upon existing analytic models limits the ability to discover new mechanisms in rich datasets. The promise of big kinematic datasets from new automated labeling methods motivates the aim for a general, data-driven paradigm to identify templates in motion data. We present a framework for identifying these dynamic posture principles grounded in a local model of the hypothesized template-anchor dynamics, enabling generalization to a range of both periodic and transient locomotion tasks. Using interpretable models of both continuous and hybrid template-anchor systems, we find that our approach is capable of effectively identifying template submanifolds with a surprisingly wide range of geometric and dynamical properties, and is robust to measurement noise and parameter errors. |
Thursday, March 9, 2023 9:36AM - 9:48AM |
S10.00007: EMBUR (EMerita Burrowing Robot): A Robophysical Exploration of Mole Crab Burrowing Laura K Treers, Benjamin McInroe, Robert J Full, Hannah S Stuart Despite the ubiquity of granular substrates like sands and soils in the natural environment, few robots are capable of burrowing vertically under their own self-weight, and even fewer can do so using legs. In this work, we discuss prior research on a novel mole crab-inspired robot– EMBUR (EMerita Burrowing Robot) – and how it has been used as a tool for robophysical exploration of legged burrowing. EMBUR’s design was guided by an initial study of the Pacific mole crab, Emerita analoga. E. analoga has five leg pairs, which on the robot are functionally simplified into two counter rotating leg pairs. The mole crabs also employ a complex leg trajectory which reduces resistive force during part of their cyclical stroke. Similarly, the legs of EMBUR are flexible and can extend and retract to create anisotropic force response. We observe parallels between the robot’s and animals’ body pitch during intrusion, as well as their spatial burrowing trajectories, suggesting that EMBUR and E. analoga employ similar excavative mechanisms for burrowing. We also explore the applications of Granular Resistive Force Theory, or RFT, to the design and implementation of EMBUR. Through parametric studies, we show that RFT can predict leg geometries and behaviors which maximize desired parameters of interest. Current research focuses on understanding discrepancies between RFT predictions and observed robot behavior, and improving the robot’s control strategies using insights from more advanced modeling techniques. |
Thursday, March 9, 2023 9:48AM - 10:00AM |
S10.00008: Hopping on Deformable Terrain with State-Based Switching: Dynamics and Implications for Robot Design Daniel J Lynch, Sean Gart, Jason Pusey, Kevin M Lynch, Paul B Umbanhowar Legged robot locomotion is hindered by a mismatch between applications where legs can outperform wheels or treads, most of which feature deformable substrates, and existing tools for planning and control, most of which assume flat, rigid substrates. While locomotion on any soft substrate can be challenging (e.g., maintaining upright posture, coordinating multiple legs), we focus on vertically-constrained single-leg hopping because it is arguably the simplest setting for studying the interplay between the depth-dependent substrate yield threshold and the forces applied during stance—a fundamental feature of legged locomotion on deformable terrain. We derive a Poincaré map that captures the hop-to-hop energy dynamics of a monopod hopping on deformable terrain. The map reveals complex boundaries in the design-control parameter space between periodic and decaying gaits, as well as families of transient responses and basins of attraction associated with each gait. These results emphasize the value of long, compliant legs and large, lightweight feet and indicate how the hop-to-hop energy dynamics can be leveraged to estimate terrain properties and to plan transitions between gaits in order to maximize efficiency, convergence rate, and robustness to terrain uncertainty. |
Thursday, March 9, 2023 10:00AM - 10:12AM |
S10.00009: Robotic morphing in the aquatic-to-terrestrial transition Robert Baines Amphibious robots are valuable platforms for studying the physics of locomotion in the aquatic-to-terrestrial transition. Waves, currents, rocks and vegetation, as well as dynamic flows of fluidized sediment, are physical phenomena a robotic system may encounter in this transition. We recently introduced Amphibious Robotic Turtle (ART), a turtle-inspired robot that adapts for locomotion in its environment via limbs that morph between specialized terrestrial and aquatic shapes and a variety of gaits. ART can swim underwater or at the surface, walk over different terrain, and transition between water and land. We investigated the confluence of robot gait, limb shape, and the environmental medium and found that, by adapting its shape and gait, ART can locomote in different environments with comparable, or in some cases, even better cost of transport than exclusively uni-modal robots. Generalized questions arose from these studies about how and when a robot should change its shape and behavior if transitioning between environments. Future work will investigate these questions through experiments with ART in a newly built aquatic-to-terrestrial transition tank featuring controllable waves, current, incline, substrate, and temperature. |
Thursday, March 9, 2023 10:12AM - 10:24AM |
S10.00010: Flipper-based locomotion on muddy substrates Boyuan Huang, Shipeng Liu, Jake Futterman, Feifei Qian Muddy substrates exist widely in nature, from wetlands to nearshore environments. The mechanical behaviors of the mud can vary significantly due to subtle changes in grain size, organic matter, and water content. Due to the limited understanding of the mechanical responses of mud, terrestrial robots still struggle to robustly cope with this wide variation. In this study, using a turtle-inspired robot, we investigated how the performance of flipper-based locomotion depends on mud properties. To systematically vary mud properties, we created synthetic mud with precisely-controlled ratios of sand, clay, and water. Tracking data suggested that in high clay ratio mixtures, the robot's step length first increased from low to intermediate water content, then decreased as water content further increased to near saturation. At low water content, the reduced step length was primarily a result of flipper extraction failures: the cohesive mud mixture could stick to the bottom of the flipper, preventing the flipper from fully lifting off the mud, and resulting in an oscillation in the robot center-of-mass (CoM) along the fore-aft direction within a step. At high water content, the reduced step length was primarily a result of the reduced yield stress, where the mixture yielded upon flipper shear, resulting in a flipper slip. |
Thursday, March 9, 2023 10:24AM - 10:36AM |
S10.00011: Biological and robophysical experiments of terrestrial fish locomotion on mud of controlled, variable strength Divya Ramesh, Gargi Sadalgekar, Qiyuan Fu, Zach Souders, Jack Rao, Chen Li Many amphibious fishes can make forays onto land. The water-land interface often has wet deformable substrate like mud and wet sand/gravel, whose strength varies with water content, which challenges locomotion. Most previous studies of fish terrestrial locomotion focused on kinematics, muscle control, and functional morphology. To fully understand how these locomotor features interact with the environment to permit performance, we need to further quantify the interaction with wet deformable substrates. Here, we used controlled mud as a model wet deformable substrate and developed methods to prepare mud into spatially uniform states and temporally stable states and tools to characterize its strength. We studied the Atlantic mudskipper (Periophthalmus barbarus) moving on a stronger and a weaker mud. The animal performed similar “crutching” walks on mud of both strengths, with only slightly slower speed on thinner mud (from 0.4 ± 0.12 to 0.31 ± 0.13 body length/s, P < 0.05, ANOVA). However, it jumped more frequently (from 1.4 ± 0.97 to 3.2 ± 1.6 times per minute, P < 0.05, ANOVA). This may be to mitigate the higher stickiness of the weaker mud. For more repeatable and controlled experiments, we created a robophysical model. The robot has two fins that rotate in phase to raise the body and “crutch” forward on mud like the mudskipper. We are currently testing the animal on even weaker and stronger mud and refining the robot for systematic experiments on varying mud strengths with varying fin morphology and kinematics. |
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. |
© 2024 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