Bulletin of the American Physical Society
APS March Meeting 2017
Volume 62, Number 4
Monday–Friday, March 13–17, 2017; New Orleans, Louisiana
Session X12: Robophysics IFocus
|
Hide Abstracts |
Sponsoring Units: DBIO GSNP GSOFT Chair: Daniel Goldman, GeorgiaTech Room: 271 |
Friday, March 17, 2017 8:00AM - 8:12AM |
X12.00001: Mixed reality framework for collective motion patterns of swarms with delay coupling Klementyna Szwaykowska, Ira Schwartz The formation of coherent patterns in swarms of interacting self-propelled autonomous agents is an important subject for many applications within the field of distributed robotic systems. However, there are significant logistical challenges associated with testing fully distributed systems in real-world settings. In this paper, we provide a rigorous theoretical justification for the use of mixed-reality experiments as a stepping stone to fully physical testing of distributed robotic systems. We also model and experimentally realize a mixed-reality large-scale swarm of delay-coupled agents. Our analyses, assuming agents communicating over an Erdos-Renyi network, demonstrate the existence of stable coherent patterns that can be achieved only with delay coupling and that are robust to decreasing network connectivity and heterogeneity in agent dynamics. We show how the bifurcation structure for emergence of different patterns changes with heterogeneity in agent acceleration capabilities and limited connectivity in the network as a function of coupling strength and delay. Our results are verified through simulation as well as preliminary experimental results of delay-induced pattern formation in a mixed-reality swarm. [Preview Abstract] |
Friday, March 17, 2017 8:12AM - 8:24AM |
X12.00002: Locomotion in a planar ideal fluid by a singly actuated elastic body Scott Kelly, Rodrigo Abrajan-Guerrero An aquatic vehicle with a single internal degree of freedom can propel itself by exploiting symmetry-breaking phenomena like vortex shedding, but the manipulation of added-mass effects to achieve locomotion in an ideal fluid --- essentially exploiting rather than breaking finite- and infinite-dimensional symmetries --- requires a swimming body to execute changes over time in at least two independent shape parameters. Such parameters may be under direct control, and prior work has addressed the design of optimal gaits for swimmers in ideal fluids under this assumption, but may also evolve dynamically as a result of partial actuation and body elasticity. This talk will describe the planar locomotion of a singly actuated jointed robot exploiting limit cycles arising in its internal shape as a result of periodic actuation. [Preview Abstract] |
Friday, March 17, 2017 8:24AM - 8:36AM |
X12.00003: Propulsion of helical flagella near boundaries Bruce Rodenborn, Grant Giesbrecht, Katha Ni, Isaac Vock The presence of nearby boundaries is known to have dramatic effects on the swimming behavior of microorganisms because of the no-slip condition at the boundary. Microorganisms that use a helical flagellum experience forces both along the axis of the helix and in the direction perpendicular to the axis. These low Reynolds number boundary effects have primarily been studied using live bacteria and using numerical simulations. However, small scale measurements give limited information about the forces and torques on the microorganisms. Furthermore, numerical studies are approximate because they have generally used Stokeslet-based simulations with image Stokeslets to represent the effects of the boundaries. Instead, we directly measure the propulsion of macroscopic helical flagella with diameter $\approx 12$mm using a fluid with viscosity $10^5$ times that of water to ensure the Reynolds number in the experiments is much less than unity, just as for bacteria. We measure the parallel and perpendicular forces as a function of boundary distance to determine the nonzero elements of the propulsive matrix for axial rotation near a boundary. We then compare our results to the theory and simulations of Lauga et al. (Biophys. J. 2006) and to biological measurements. [Preview Abstract] |
Friday, March 17, 2017 8:36AM - 8:48AM |
X12.00004: A coin vibrational motor swimming at low Reynolds number Alice Quillen, Hesam Askari, Doug Kelley, Tamar Friedmann, Patrick Oakes Low-cost coin vibrational motors, used in haptic feedback, exhibit rotational internal motion inside a rigid case. Because the motor case motion exhibits rotational symmetry, when placed into a fluid such as glycerin, the motor does not swim even though its oscillatory motions induce steady streaming in the fluid. However, a piece of rubber foam stuck to the curved case and giving the motor neutral buoyancy also breaks the rotational symmetry allowing it to swim. We measured a 1 cm diameter coin vibrational motor swimming in glycerin at a speed of a body length in 3 seconds or at 3 mm/s. The swim speed puts the vibrational motor in a low Reynolds number regime similar to bacterial motility, but because of the oscillations of the motor it is not analogous to biological organisms. Rather the swimming vibrational motor may inspire small inexpensive robotic swimmers that are robust as they contain no external moving parts. A time dependent Stokes equation planar sheet model suggests that the swim speed depends on a steady streaming velocity that depends on the streaming Reynolds number. [Preview Abstract] |
Friday, March 17, 2017 8:48AM - 9:00AM |
X12.00005: Tuning Bacterial Hydrodynamics with Magnetic Fields: A Path to Bacterial Robotics Christopher Pierce, Eric Mumper, Jack Brangham, Hiran Wijesinghe, Stephen Lower, Brian Lower, Fengyuan Yang, Ratnasingham Sooryakumar Magnetotactic Bacteria (MTB) are a group of motile prokaryotes that synthesize chains of lipid-bound, magnetic nano-particles. In this study, the innate magnetism of these flagellated swimmers is exploited to explore their hydrodynamics near confining surfaces, using the magnetic field as a tuning parameter. With weak (Gauss), uniform, external, magnetic ?elds and the field gradients arising from micro-magnetic surface patterns, the relative strength of hydrodynamic, magnetic and ?agellar force components is tuned through magnetic control of the bacteria's orientation and position. In addition to direct measurement of several hydrodynamic quantities related to the motility of individual cells, their tunable dynamics reveal a number of novel, highly controllable swimming behaviors with potential value in micro-robotics applications. Specifically, the experiments permit the MTB cells to be directed along parallel or divergent trajectories, suppress their flagellar forces through magnetic means, and induce transitions between planar, circulating trajectories and drifting, vertically oriented ``top-like'' motion. The implications of the work for fundamental hydrodynamics research as well as bacterially driven robotics applications will be discussed. [Preview Abstract] |
Friday, March 17, 2017 9:00AM - 9:12AM |
X12.00006: Soap-bubble Optimization of Gaits Suresh Ramasamy, Ross Hatton We present a geometric gait optimizer that applies Lie bracket theory to identify optimal cost-of-transport (displacement divided by effort) gaits. This optimizer builds on our previous work, where we have shown that for drag-dominated systems, the efficiency of a gait corresponds to a ratio between ``metric-weighted” perimeter length of the cycle and the area integral of the Lie bracket it encloses. In this work, we encode this geometric insight into a variational gait optimizer. For a system with two shape variables, the dynamics of this optimizer are similar to the dynamics of a soap bubble, with the Lie bracket providing internal pressure which causes the boundary of the bubble to expand, the metric-weighted path length providing surface tension constraining the growth of the soap bubble, and a pace-balancing term corresponding to the concentration gradient that evenly distributes soap across the surface of the bubble. In systems with three shape variables, the dynamics are more akin to a windsock, capturing maximum flux through a loop. The variational form of the optimizer allows us to extend it to higher dimensional shape spaces beyond these physical analogies. [Preview Abstract] |
Friday, March 17, 2017 9:12AM - 9:24AM |
X12.00007: Gait Costs and Geodesic Curvature Hossein Faraji, Ross Hatton In a drag-dominated environment, the effort required to change shape can be modeled as the pathlength of the trajectory through the shape space, with a Riemannian distance metric induced by the energy dissipated through friction. For an inertia-dominated system, the Riemannian metric is induced by the inertia matrix for the system, and effort corresponds to accelerations that change the system's momentum. Optimizing costs of trajectories has two fundamental cases: The first case is finding a minimum cost trajectory between two points, where the shortest and least curving trajectory is a geodesic of the drag or inertial tensor. The second case is minimizing the cost to connect three or more waypoints. For a drag-based system, this trajectory is a collection of geodesic segments, but for an inertial system, it is instead a geodesic spline that avoids sharp corners in the path. A geodesic spline with constant speed minimizes geodesic curvature squared. With non-constant speed, it minimizes squared magnitude of the acceleration, comprising tangential acceleration and speed-weighted curvature). We apply a variational approach to find geodesic splines to generate optimal gaits on serpenoid and three-link inertial snakes, and optimal swinging trajectories for legged robots. [Preview Abstract] |
Friday, March 17, 2017 9:24AM - 9:36AM |
X12.00008: Phase Helps Find Geometrically Optimal Gaits Shai Revzen, Ross Hatton Geometric motion planning describes motions of animals and machines governed by $\dot g = g A(q) \dot q$ -- a connection $A(\cdot)$ relating shape $q$ and shape velocity $\dot q$ to body frame velocity $g^{-1} \dot g \in \mathrm{se}(3)$. Measuring the entire connection over a multidimensional $q$ is often unfeasible with current experimental methods. We show how using a phase estimator can make tractable measuring the local structure of the connection surrounding a periodic motion $q(\varphi)$ driven by a phase $\varphi\in \mathsf{S}^1$. This approach reduces the complexity of the estimation problem by a factor of $\dim q$. The results suggest that phase estimation can be combined with geometric optimization into an iterative gait optimization algorithm usable on experimental systems, or alternatively, to allow the geometric optimality of an observed gait to be detected. [Preview Abstract] |
Friday, March 17, 2017 9:36AM - 9:48AM |
X12.00009: Mimicking Muscle Nonlinear Force Generation using Electromagnetic Motors José Alvarado, Anette Hosoi Animals routinely perform a wide range of mechanical tasks, including locomotion, and continue to inspire solutions in engineering applications. Yet despite numerous technological advances, robotic locomotion lags behind that of animals in terms of versatility and energy economy. One reason for this performance gap lies in actuation: electromagnetic motors are common actuators in engineered systems, whereas animals primarily use muscle. Researchers have long modeled muscle with a nonlinear force-velocity relationship, in contrast to motors’ linear behavior. Existing theoretical studies have predicted advantages to nonlinear force generation, including energy economy, stability, and simplified controls. Yet these advantages are difficult to verify experimentally because the force-velocity curve of intact muscle cannot be made linear. Here we establish a physical model system of muscle nonlinearity by programming an electromagnetic motor to exhibit linear and nonlinear behavior. Preliminary experimental and theoretical results show that for the simple task of lifting a weight against gravity, muscle-like nonlinearity merely reduces work output. We anticipate that for more complex mechanical tasks, muscle’s nonlinear properties could be mechanically advantageous. [Preview Abstract] |
Friday, March 17, 2017 9:48AM - 10:00AM |
X12.00010: Task driven optimal leg trajectories in insect-scale legged microrobots Neel Doshi, Benjamin Goldberg, Kaushik Jayaram, Robert Wood Origami inspired layered manufacturing techniques and 3D-printing have enabled the development of highly articulated legged robots at the insect-scale, including the 1.43g Harvard Ambulatory MicroRobot (HAMR). Research on these platforms has expanded its focus from manufacturing aspects to include design optimization and control for application-driven tasks. Consequently, the choice of gait selection, body morphology, leg trajectory, foot design, etc. have become areas of active research. HAMR has two controlled degrees-of-freedom per leg, making it an ideal candidate for exploring leg trajectory. We will discuss our work towards optimizing HAMR’s leg trajectories for two different tasks: climbing using electroadhesives and level ground running (5-10 BL/s). These tasks demonstrate the ability of single platform to adapt to vastly different locomotive scenarios: quasi-static climbing with controlled ground contact, and dynamic running with un-controlled ground contact. We will utilize trajectory optimization methods informed by existing models and experimental studies to determine leg trajectories for each task. We also plan to discuss how task specifications and choice of objective function have contributed to the shape of these optimal leg trajectories. [Preview Abstract] |
Friday, March 17, 2017 10:00AM - 10:12AM |
X12.00011: Synchronization of RLC oscillators Helene Nguewou-Hyousse, Derek Paley Caterpillars' motion, although not fast or efficient, has some advantages: They are flexible and can move through complex terrains, and there is no difference between horizontal and vertical locomotion. For these reasons, caterpillar-inspired robots may be safe and reliable, with applications in search and rescue. This talk introduces an approach to modeling the locomotion dynamics of a caterpillar using a network of RLC circuits connected through spring-like and damping local interactions. Graph theory is used as a tool to describe a bidirectional network of RLC oscillators with and without a leader (\textit{deC}entralized Pattern Generator). The synchronization of the system is verified theoretically and experimentally, and the stability is studied using matrix theory and phase model analysis. [Preview Abstract] |
Friday, March 17, 2017 10:12AM - 10:24AM |
X12.00012: Dynamic traversal of high bumps and large gaps by a small legged robot Sean Gart, Nastasia Winey, Rafael de la Tijera Obert, Chen Li Small animals encounter and negotiate diverse obstacles comparable in size or larger than themselves. In recent experiments, we found that cockroaches can dynamically traverse bumps up to 4 times hip height and gaps up to 1 body length. To better understand the physics that governs these locomotor transitions, we studied a small six-legged robot negotiating high bumps and large gaps and compared it to animal observations. We found that the robot was able to traverse bumps as large as 1 hip height and gaps as wide as 0.5 body length. For the bump, the robot often climbed over to traverse when initial body yaw was small, but was often deflected laterally and failed to traverse when initial body yaw was large. A simple locomotion energy landscape model explained these observations. For the gap, traversal probability decreased with gap width, which was well explained by a simple Lagrangian model of a forward-moving rigid body falling over the gap edge. For both the bump and the gap, animal performance far exceeded that of the robot, likely due to their relatively higher running speeds and larger rotational oscillations prior to and during obstacle traversal. Differences between animal and robot obstacle negotiation behaviors revealed that animals used active strategies to overcome potential energy barriers. [Preview Abstract] |
Friday, March 17, 2017 10:24AM - 11:00AM |
X12.00013: Passive dynamics is a good basis for robot design and control, not! Invited Speaker: Andy Ruina Many airplanes can, or nearly can, glide stably without control. So, it seems natural that the first successful powered flight followed from mastery of gliding. Many bicycles can, or nearly can, balance themselves when in motion. Bicycle design seems to have evolved to gain this feature. Also, we can make toys and 'robots' that, like a stable glider or coasting bicycle, stably walk without motors or control in a remarkably human-like way. Again, it seems to make sense to use `passive-dynamics' as a core for developing the control of walking robots and to gain understanding of the control of walking people. That's what I used to think. But, so far, this has not led to robust walking robots. What about human evolution? We didn't evolve dynamic bodies and then learn to control them. Rather, people had elaborate control systems way back when we were fish and even worms. However: if control is paramount, why is it that uncontrolled passive-dynamic walkers walk so much like humans? It seems that energy optimal, yet robust, control, perhaps a proxy for evolutionary development, arrives at solutions that have some features in common with passive-dynamics. Rather than thinking of good powered walking as passive walking with a small amount of control added, I now think of good powered walking, human or robotic, as highly controlled, while optimized for, in part, minimal actuator use. Thus, much of the motor effort, always at the ready, is usually titrated out. [Preview Abstract] |
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