Bulletin of the American Physical Society
APS March Meeting 2020
Volume 65, Number 1
Monday–Friday, March 2–6, 2020; Denver, Colorado
Session S22: Robophysics IFocus Undergrad Friendly
|
Hide Abstracts |
Sponsoring Units: DBIO Chair: Daniel Goldman, Georgia Institute of Technology Room: 303 |
Thursday, March 5, 2020 11:15AM - 11:51AM |
S22.00001: Adaptive robot morphology and behavioral control policy through responsive composite materials Invited Speaker: Rebecca Kramer-Bottiglio Robots generally excel at specific tasks in structured environments, but lack the versatility and adaptability required to interact with the real world. In this talk, I will introduce several soft robot platforms that can adapt both their shape and behavior to accommodate different tasks or environments. While actuation alone may temporarily change a robot’s shape, I will show that the ability to change resting shape is an advantage of soft robots, and that shape control can offset computationally-expensive tasks such as gait changes and obstacle avoidance. The focus will be on the responsive composites that enable these functional capabilities, including variable stiffness materials for move-and-hold shapeshifting applications, switchable stretchability materials for adaptive trajectories of volumetrically-expanding actuators, and stretchable circuitry for material-embedded computation. |
Thursday, March 5, 2020 11:51AM - 12:03PM |
S22.00002: Robots and animals transition from less to more favorable modes to traverse obstacles Ratan Sadanand Othayoth Mullankandy, George Thoms, Chen Li Legged robots still struggle to robustly traverse obstacles in complex terrain using physical interaction, which animals do at ease. Recent studies revealed that cockroaches use body-beam interaction to traverse grass-like beam obstacles, probabilistically transitioning between locomotor modes. To understand what governs the direction of locomotor transitions between modes, we developed and studied a robophysical model traversing beams of variable stiffness. We discovered that, when viewed on a potential energy landscape resulting from body-beam interaction, different locomotor modes (system state trajectories) were attracted to different local minima basins. Regardless of beam stiffness, the system was always more likely to transition from less to more favorable modes (i.e., from lower to higher energy local minima) on the landscape, as long as kinetic energy fluctuation from oscillatory self-propulsion is sufficient to overcome the potential energy barrier in between. We made similar observations in cockroaches, despite their active behaviors. These physical principles are surprisingly similar to microscopic systems (e.g., protein-folding transitions on a free energy landscape). Our study is a step in establishing energy landscapes for locomotor transitions in complex terrain. |
Thursday, March 5, 2020 12:03PM - 12:15PM |
S22.00003: Active adjustments help cockroaches traverse obstacles by lowering potential energy barrier Yaqing Wang, Ratan Sadanand Othayoth Mullankandy, Chen Li Robots predominantly navigate complex environments by using geometric models to avoid physical contact with obstacles. By contrast, animals easily traverse obstacles by using physical interaction controlled by active feedback. Our previous study revealed that cockroaches made active adjustments to traverse grass-like beam obstacles. As the animal pushed against the beams, its body pitched up. Then, the animal often quickly rolled into the gap between beams to traverse. In this process, the animal’s head often flexed repeatedly, and one hind leg pushed against the ground while the other retracted. Here, we used physics modeling and robot experiments to understand the function of such active adjustments. We calculated system potential energy (resulting from gravity and beam elastic bending) as a function of head flexion. When the animal was in contact with the beams, head flexion lowered its potential energy barrier to transition from pitch to roll mode (by nearly half for the observed head flexion of 20°). We hypothesize that the animal uses proprioception (internal position and force sensing) to detect changes in beam resistive forces, which reflect the reducing transition barrier. We are developing a robotic physical model instrumented with force sensors to test this hypothesis. |
Thursday, March 5, 2020 12:15PM - 12:27PM |
S22.00004: A template model reveals self-righting mechanism of a winged robot Qihan Xuan, Chen Li Animals and robots must self-right after flipping-over on the ground. Winged insects like many cockroaches push wings against the ground while flailing legs to generate lateral perturbation to self-right. Our previous study of a cockroach-inspired robot revealed the importance of good coordination (measured by phase) between wing pushing and leg flailing. Here, we elucidate the mechanism of phase dependence by developing a template model, validated against multi-body dynamics simulation. The model consists of a point mass body rotating in the sagittal plane, two massless wings, and a flailing leg with a point mass at its end. With modest wing opening and leg flailing, the model struggled to self-right, and successful righting relied on good coordination of the wings and leg. We used the model to calculate the potential energy barrier the body must overcome, mechanical energy input by the wings and leg, and energy dissipation due to collision and friction. With good phases, mechanical energy accumulation exceeded potential energy barrier; with bad phases, it did not do so. We used the template model to predict an optimal coordination strategy to increase self-righting probability. Our study highlighted the importance of appendage coordination in strenuous locomotor transitions. |
Thursday, March 5, 2020 12:27PM - 12:39PM |
S22.00005: Testing the effect of scaling on microrobot locomotion performance Kaushik Jayaram, Robert Wood The effects of scaling on locomotion performance has fascinated biologists, physicists and engineers alike. In particular, roboticists have exploited dynamic scaling to build systems at sizes varying from a tens of centimeters to several meters. Thanks to recent advances in manufacturing, insect-scale robotics is currently an exciting research direction and holds the promise of tremendous impact in areas of search-and-rescue and high-value asset inspection. However, most robots at this scale have simplified morphology and can only demonstrate basic mobility. Here, we present the newly designed HAMR-Jr, a 22.5mm, 320mg quadrupedal microrobot. With eight independently actuated degrees of freedom, HAMR-Jr is, to our knowledge, the most mechanically dexterous legged robot at its scale and is capable of high-speed locomotion (13.91 bodylengths/s) at a variety of stride frequency (1-200Hz) using multiple gaits. We achieved this using a design and fabrication process that is flexible, allowing us to exploit and implement the physics of scaling with minimum changes to our workflow. We further characterized HAMR-Jr’s open-loop locomotion and compared it with the larger scale HAMR-VI microrobot to demonstrate the effectiveness of scaling laws in predicting running performance. |
Thursday, March 5, 2020 12:39PM - 12:51PM |
S22.00006: Tail control decreases jamming of a mechanics-dominated legged robot on rough terrain Daniel Soto, Daniel I Goldman To develop robot locomotion principles which leverage mechanics to simplify control, we study a minimally aware quadrupedal RHex-type robot (~3kg, 20 cm long) walking on a rough terrain composed of blocks with Gaussian height distribution with mean and variance approximately leg length and one-quarter leg length respectively. This terrain resulted in failure modes due to limb jamming in protruding regions of the terrain. Performance was insensitive to gait symmetries (phasing and duty factor of limbs). To reduce jamming probability while maintaining the focus on minimal control and mechanics, we added a motor driven 10 cm long tail whose pitch and yaw actuation was uncoupled to limb dynamics. We experimented with a variety of open loop behaviors such as tapping, pulsing, and maintaining a constant angle with the body. We also developed a closed loop tapping scheme that triggered on IMU measured pitch and roll of the robot. Jamming probability could be reduced to near zero for open and closed loop schemes. The appropriate tail behaviors helped maintain either high or low “robot temperature” (variance of instantaneous velocity) relative to tail-less robot temperature. |
Thursday, March 5, 2020 12:51PM - 1:03PM |
S22.00007: Modulation of robot orientation by exploiting leg-obstacle collisions through successive choices of gait Feifei Qian, Yongxin Guo, Anmol Kathail, Daniel Koditschek An environment can provide different affordances for locomotion based on a locomotor’s actions. Our study aims to explore how multi-legged robots can use various gaits to exploit the affordance of obstacle collisions to negotiate complex environments in different ways. |
Thursday, March 5, 2020 1:03PM - 1:15PM |
S22.00008: A systematic approach to creating terrain-capable hybrid soft/hard myriapod robots Yasemin Ozkan-Aydin, Enes Aydin, Baxi Chong, Daniel I Goldman Multi-legged animals such as centipedes move effectively in diverse terrain. Their flexible body and limbs allow them to adapt their shape to the environment and traverse obstacles. To examine how the variation in body/limb forms of legged animals affects the mechanics of terrestrial locomotion, we built a multi-legged hybrid (soft/rigid) robot that has 8 segments (9 cm/segment), each with two out of phase legs. Back elements and limbs are driven by servo motors. We systematically tested the gait patterns with different leg contacts and body undulation provided by geometric mechanics on various environments including flat ground, obstacle course, stair climbing, and unstructured natural terrain. On flat ground, the robot with completely rigid parts moved in the same way as theoretically predicted gaits. As the complexity of the environment increases, adding flexibility into the robot body parts (legs, body joints, etc.) improved the locomotion performance by either reducing the effects of environmental disturbances or increasing stability. |
Thursday, March 5, 2020 1:15PM - 1:27PM |
S22.00009: Basilisk Lizard Bipedal Locomotion on Sand, Mud, and Water: An Insight to Future Morphing Amphibious Robots Hosain Bagheri, Marcus Gambatese, David Lentink, Hamidreza Marvi With the astonishing development in robotics, one area that still remains an open challenge is their physical interaction with the outside world, beyond that of factory floors. In order for robotic devices to integrate and conquer obstacles introduced by the surrounding world, they will need to be able to traverse on and through diverse and alternating mediums and terrains. To address this desire of adaptive multi-modal locomotion, we look to the basilisk lizard as a form of inspiration and means of understanding how it can effectively run and transition from land to water, and vice versa. By shedding light on both active (i.e. gait, body dynamics) and passive (i.e. tail swing, toe fringes) mechanisms used by the animal, more competent robotic systems can be developed for the interaction and exploration of both terrestrial and aquatic environments. Notably, passive mechanisms do not require input power, sensing, or controls for operation. Through conducting animal experiments in both terrestrial and aquatic environments (and the transition between the two), the hierarchical physics of animal interaction with complex environments can be extracted and channeled towards the development of a bio-inspired amphibious robot for planetary exploration and search-and-rescue applications. |
Thursday, March 5, 2020 1:27PM - 1:39PM |
S22.00010: Hierarchical control in sea star inspired locomotion Sina Heydari, Theodora Po, Matthew McHenry, Eva Kanso There is a growing effort to understand decentralized control mechanisms, particularly in application to robotic systems with distributed sensors and actuators. Sea stars, being equipped with hundreds of tube feet, are an ideal model system for studying decentralized sensing and actuation. The activity of the tube feet is orchestrated by a nerve net that is distributed throughout the body; there is no central brain. We developed mathematical models of the biomechanics of the tube feet and the sea star body. We then formulated hierarchical control laws that capture salient features of the sea star nervous system. Namely, at the component level, the individual tube feet follow a state-dependent feedback controller. At the system level, a directionality command is communicated to all tube feet. We studied the locomotion gaits afforded by this control model. We find that these minimally-coupled tube feet coordinate to generate robust forward locomotion on different terrains. Our model also predicts different gait transitions consistent with our experiments performed on Protoreaster nodosus. These findings offer a new paradigm for walking using soft actuators, with potential applications to autonomous robotic systems. |
Thursday, March 5, 2020 1:39PM - 1:51PM |
S22.00011: Hopping with Elastic Restitution Is More Difficult Than It Seems Taylor McLaughlin, Marion Anderson, Shai Revzen Bouncing rubber balls and hopping on pogo sticks might suggest that building a robot that hops with a significant fraction of elastic energy restitution would be straightforward. Using spring steel legs with 99% restitution individually, we built a three-legged hopper. The hopper achieved a minute 11.2% restitution when using the springs in series with the actuators. We present five generations of three-legged hoppers: four have actuators in series with springs. Three different types of ground contact were tested. Our best design so far uses actuators in parallel with the springs, and achieves 49+/-5% restitution. This work will be of use to those interested in using elastic energy storage in legged locomotion. |
Thursday, March 5, 2020 1:51PM - 2:03PM |
S22.00012: Recovery of Behaviors of Robots without Dynamics George Council, Shai Revzen For robots to function without humans on hand to fix them, they must be able to compensate for malfunctions that inhibit their ability to move all actuators. |
Thursday, March 5, 2020 2:03PM - 2:15PM |
S22.00013: Modeling of the Primary Animal Gaits by Coupled Identical Raleigh-Van der Pol Oscillators Masoud Asadi-zeydabadi, Mozhdeh Saffari-Parizi, Randall Tagg Mathematical models based on coupled oscillators successfully describe animal gaits. The dynamics of individual oscillators (internal dynamics) in the network is nonlinear and must be at least two-dimensional to have a Hopf bifurcation. One example of this type of dynamics is a network of Rayleigh - Van der Pol oscillators which is used to produce primary gaits such as walk, trot, pace and bound. When these oscillators couple together, they produce patterns of relative phases that correspond to the different animal gaits. The patterns emerge through bifurcations as parameters describing the coupling are varied. At least eight cells are required to model the quadruped gaits. |
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