Bulletin of the American Physical Society
APS March Meeting 2020
Volume 65, Number 1
Monday–Friday, March 2–6, 2020; Denver, Colorado
Session X22: Robophysics IV and Animal Behavior |
Hide Abstracts |
Sponsoring Units: DBIO Chair: Kaushik Jayaram, University of Colorado, Boulder Room: 303 |
Friday, March 6, 2020 11:15AM - 11:27AM |
X22.00001: Controlling robot dynamics via environmental deformations Hussain Gynai, Shengkai Li, Yasemin Ozkan-Aydin, Camila Dominguez, Enes Aydin, Pablo Laguna, Daniel I Goldman In an effort to construct a robophysical analog gravity system, we previously studied the dynamics of a 200-gram differential wheeled robot driving on a deformable spandex membrane (d=2.4m) with a static central depression (Li et al., 2019). The system displays rich dynamics, including precessing orbits reminiscent of those in general relativity. Remarkably, the vehicle-membrane system can be mapped to the dynamics of a test particle in a fiducial spacetime. To take the next step and study how the vehicle dynamics can be manipulated via dynamic non-local changes in membrane curvature, we developed an automated gantry system that can control the position of a spherical object which creates a local deformation on the membrane. From experiments implementing a closed, circular object trajectory, the translating depression can capture the locomoting car from specific initial conditions, depending on initial orientation and position of the car and object trajectories. We explore how feedback from internally based measurements of the vehicle’s tilt and acceleration can enhance capture rates. |
Friday, March 6, 2020 11:27AM - 11:39AM |
X22.00002: Modeling and simulation of complex dynamic musculoskeletal architectures Xiaotian Zhang, Fan Kiat Chan, Tejaswin Parthasarathy, Mattia Gazzola Natural creatures, from fish and cephalopods to snakes and birds, combine neural control, sensory feedback and compliant mechanics to effectively operate across dynamic, uncertain environments. In order to facilitate the understanding of the biophysical mechanisms at play and to streamline their potential use in engineering applications, we present here a versatile and robust numerical approach to the simulation of musculoskeletal architectures. It relies on the assembly of heterogenous, active and passive Cosserat rods into dynamic structures that model bones, tendons, ligaments, fibers and muscle connectivity. We demonstrate its utility in a range of problems involving biological and soft robotic scenarios across scales and environments: from the engineering of millimeter-long bio-hybrid robots to the synthesis and detailed reconstruction of complex musculoskeletal systems. The versatility of our methodology offers a powerful framework to aid forward and inverse bioengineering designs as well as fundamental discovery in the functioning of living organisms. |
Friday, March 6, 2020 11:39AM - 11:51AM |
X22.00003: Chrono: A multi-physics engine for simulation of robophysical systems Milad Rakhsha, Radu Serban, Dan Negrut Early-stage computer analysis of robophysical systems informs critical design decisions, where systematic study of the design-space via experiment can be challenging and expensive. We present the latest capabilities of a multi-physics platform, called Chrono, that allows for computer modeling and simulation of such systems. The term multi-physics is used herein as a broad umbrella for rigid body dynamics with frictional-contact, flexible body dynamics, fluid dynamics, and fluid-solid interaction. This paradigm is critical in many robophysical systems since locomotion of rigid/flexible robots in conjunction with rigid, granular, and fluid substrates is ubiquitous. The simulation platform supports applications that range from locomotion of robotic agents over non-trivial geometries and on gravel, to underwater robotics. Furthermore, we describe how these simulations can be used as episodes of deep reinforcement learning models for training smart agents who can learn efficient locomotion pathways through machine learning. This is important because (i) fine-tuning the optimum design parameters can be laborious, and (ii) training agents in a virtual environment instead of the real world reduces the time and risks of the operation. |
Friday, March 6, 2020 11:51AM - 12:03PM |
X22.00004: Synchrono: A simulation framework for machine learning in multi-robot applications. Asher Elmquist, Radu Serban, Dan Negrut We present here a simulation infrastructure designed to allow for safe, low-cost, and rapid development, testing, and evaluation of robot control strategies particularly in scenarios that are difficult or impossible to test in reality. The software platform provides (a) simulated dynamics to support interaction between the robots and the environment, (b) simulated sensing to provide the robots with a realistic perspective of the environment, (c) simulated communication to support inter-robot communication used for collaboration and coordination. Support of dynamics allows the exploration of physics-limited scenarios such as slip, collisions, non-linear flexibility, or interaction with fluid. Combining the dynamic simulation with virtual sensing allows us to enhance the capability of the control strategies by exploring sensing limited scenarios such as low light or inclement weather. In addition to dynamics and sensing, simulated communication in the virtual environment allows for coordination that may be used in multi-robot scenarios. These complex scenarios that include dynamics, sensing, and communication are parallelized in this simulation framework such that hundreds of variations can be performed to probe difficult or impossible to test edge-cases. |
Friday, March 6, 2020 12:03PM - 12:15PM |
X22.00005: Learning to locomote in the presence of symmetry Scott Kelly Deep reinforcement learning algorithms provide a paradigm whereby a biomorphic robot can refine a strategy for efficient locomotion based on judicious trial and error, exploiting a biologically inspired architecture for the storage of experiential knowledge. An important feature of this paradigm is its applicability to systems for which accurate mathematical models are unavailable, so that behavioral policies must be constructed directly from sensor feedback. It’s commonly the case that even when a mathematical model is unavailable for a system’s dynamics, fundamental considerations ensure that symmetries underlie these dynamics. Practical reinforcement learning in a physical setting requires a parsimonious approach to data collection and representation. This talk will discuss the use of symmetry to improve the economy with which a physical robot can learn to locomotive efficiently. |
Friday, March 6, 2020 12:15PM - 12:27PM |
X22.00006: Nonlinear distance and velocity estimation from optic flow Floris van Breugel Vision is widely used by animals and robots for position and velocity estimation. Robots typically use stereopsis or feature recognition to extract distance from image data. Stereopsis, however, requires 2 calibrated cameras positioned at a sufficiently large distance, and feature based methods require computationally expensive image processing. These limitations preclude small animals and robots from estimating distance in this manner. What alternative solutions exist? I will review a nonlinear estimation method for separating distance and velocity information from optic flow generated by a single dynamically moving camera. Unfortunately, this approach requires calculating the time derivative of optic flow, a notoriously noisy signal. To overcome this limitation, we use a new neural network optic flow estimator, flownet2, and total variation regularization methods to estimate smooth derivatives. These steps allow us to independently measure distance and velocity from a single dynamically moving camera. We propose that this approach to integrating multiple sensory modalities during dynamic motion, which we refer to as idiokinemetry, is likely a general feature of how animals perceive the world, and may inspire the development of smaller and more robust robotic systems. |
Friday, March 6, 2020 12:27PM - 12:39PM |
X22.00007: Raising head facilitates antenna cleaning in a honey bee Wei Zhang, Jianing Wu, Zhigang Wu Antennae are arguably of significance for honey bees to engage in olfaction, foraging, flying and other physiological behaviors. Contamination on the hairy antennae by dust and pollen is an ineluctable problem. Honey bees have a special structure, namely the antenna cleaner on their forelegs, to clean their antennae with. The antenna cleaner is composed of a deep notch on the ventral surface of the proximal end of the basitarsus and a large spur, which is movable, inserted at the inner angle of the distal end of the tibia. To maintain cleanliness, a honey bee grooms antennae frequently by swinging forelegs with the elaborate antenna cleaner. By high-speed imaging, we find that the honey bee raises its head while grooming antennae by forelegs. By theoretical and experimental combined study, we demonstrate that the head-raising behavior accelerates relative speed by 310% and increases friction force by 70%, which facilitates a swift and efficient grooming stroke. This behavior may inspire next-generation cleaning devices applied in microelectromechanical systems. |
Friday, March 6, 2020 12:39PM - 12:51PM |
X22.00008: Large scale quantitative phenotyping of aversive behaviour and habituation in C. elegans Luigi Feriani, Ida Barlow, Adam McDermott-Rouse, Andre Brown The nematode C. elegans has been influential in the development of the physics of behaviour, especially in methods for representing and analysing postural dynamics. To apply these methods in the context of genetic and drug screens, we need to be able to record behaviour with high resolution and throughput. We have combined robotic liquid handling, multi-camera imaging systems, and tracking software in a high-throughput pipeline that can record from about 500 samples simultaneously. Quantitative analysis of the spontaneous locomotion of C. elegans already yields a multidimensional phenotype that allows us to detect even subtle drug treatment-induced changes in its behaviour. We have added a wide-area, intense blue LED illumination to our recording system to increase the dimensionality of the explored phenotype space by both analysing the response elicited by a single stimulus, and studying simple of forms of learning such as habituation to repeated stimulations. |
Friday, March 6, 2020 12:51PM - 1:03PM |
X22.00009: Using robotics and physics to understand the evolution of novel functional morphologies Brooke Flammang, Kaelyn Gamel, Austin Garner Remoras are fishes that attach to other swimming organisms via an adhesive disc evolved from dorsal fin elements. However, the factors driving the evolution of remora disc morphology are poorly understood. Fortunately, the fundamental physics of suction and friction are mechanically conserved through time. Using a morphologically relevant bioinspired remora disc, we experimentally investigated the performance of hypothetical evolutionary intermediates. Herein, we translated fundamental biological principles into engineering design rules and show that a passive model system can autonomously achieve adhesive forces measured in live remoras in any environment. Our experimental results show that an increase in lamellar number resulted in an increase in shear adhesive performance, supporting the phylogenetic trend observed in extant remoras. The greatest pull-off forces measured for our model were on surface roughness on the order of shark skin and exceeded those measured for live remoras attached to shark skin by almost 60%. Overall, relative to fossil remoras and their closest ancestor, extant remoras exhibit a morphology indicative of selection for enhanced shear adhesive performance. |
Friday, March 6, 2020 1:03PM - 1:15PM |
X22.00010: Electrically Programmable Micro-Scale Shape Memory Devices Qingkun Liu, Wei Wang, Michael Reynolds, Marc Miskin, Michael Cao, David Muller, Paul L McEuen, Itai Cohen We demonstrate microscale shape memory devices capable of achieving micrometer bending radius with a holding time of several tens of hours. The core of these devices consists of a nanometer-thin platinum layer capped on one side by titanium. Under application of potentials in the range of 1 volt, ions oxidize the platinum, create a differential in stress between the two sides, and cause the structure to bend. Using thick panels we can localize the bending and readily create 3D shapes and patterns that can be reversibly erased and rewritten by short electrical pulses. This electrical programmability can be harnessed to make sequential and bidirectional folding. We demonstrate several reconfigurable 3D patterns, kirigami, and origami motifs. As such these micro-scaled shape memory devices enable a variety of applications in fields ranging from mechanical memory storage to microscopic robots. |
Friday, March 6, 2020 1:15PM - 1:27PM |
X22.00011: The radical pair mechanism can provide a sensitive and robust magnetic compass for birds. Shawn Strausser, Thorsten Ritz Many adult birds can travel 5,000km or more with a precision of centimeters. Earth's magnetic field provides an omnipresent source of information that aides in this navigation. The mechanism by which animals sense the Earth's magnetic field remains one of the most important problems in sensory biology. The radical pair mechanism (RPM) proposes that the Earth's magnetic field influences a chemical pathway via electron spins in a radical pair. Recently it has been shown that a so called “quantum needle” radical pair model can be extraordinarily sensitive to directional changes, thus potentially providing the basis for a highly sensitive magnetic. However, it has also been shown that many compass systems are fragile, i.e. that even small, naturally occurring, variations in parameters can abolish effects that are observed in models without considering such noise. Here, we ask what effects noise has on a quantum needle compass system. Rather than reducing the quantum needle, noise is found to either leave the needle intact or even enhance it in some systems. This suggest that a quantum needle based compass may not only be distinguished by a high sensitivity, but also by being unusually robust to noise, making it a highly optimized system for magnetoreception. |
Friday, March 6, 2020 1:27PM - 1:39PM |
X22.00012: Breathing from Underground: Diurnal Variability in the Ventilation Mechanism of Termite Mounds Saurabh Saxena, Neda Yaghoobian Termites are social insects that build massive porous mounds for creating a habitable environment for their colony. These structures exhibit intricate architectures that can effectively harness natural wind, solar energy, and colony’s metabolic heat to produce controlled microclimates in termite nests. Based on experimental studies, several mechanisms have been proposed to explain the ventilation and gas exchange process in termite mounds, such as, metabolism-driven convection, the impact of external wind, and formation of thermal gradients due to diurnal variation of the mound surface temperature. |
Friday, March 6, 2020 1:39PM - 1:51PM |
X22.00013: Specialisation and plasticity in a primitive social insect Solenn Patalano, Adolfo Alsina, Wolf Reik, Steffen Rulands Biological systems rely on an influx of energy from the environment to build and maintain complex spatio-temporal structures in noisy environments. It has recently become evident that they also have the remarkable capacity to break up and rebuild such structures, exemplified by the capability of differentiated cells to reprogram after injury. Here we use primitive societies of Polistes wasps as a model system where we experimentally perturb the nest and follow the re-establishment of the social steady state. We combine a unique experimental strategy correlating measurements across vastly different spatial scales with a theoretical approach to show that Polistes integrates antagonistic processes on different scales to simultaneously achieve plasticity and robust specialisation. We show that stability of the nest relies on epigenetic DNA modifications that suppress transcriptional noise. Such dynamics provide a general principle of how both specialization and plasticity can be achieved in biological systems. |
Friday, March 6, 2020 1:51PM - 2:03PM |
X22.00014: Hydrodynamics of sheep herds Raphael Sarfati, Marine de Marcken, Orit Peleg The collective motion of living entities, from cells to wildebeests, is a fascinating and aesthetic phenomenon which continues to intrigue physicists as well as any curious observer. Over the past decade, many compelling findings have come out of the rigorous study of systems such as bird flocks and mosquito swarms, where individuals are scattered and interact remotely. In parallel, slowly-moving aggregates of grasping individuals, such as ants and bees, have been successfully investigated from a material perspective. A third category of interest is now surfacing, in which individuals move collectively while remaining in a near-jamming state. This includes for example crowds of humans, or herds of sheep, which are the focus of this presentation. By employing optical flow methods to measure the velocity field of dense sheep herds, we construct a hydrodynamics framework to study the collective motion of living systems flowing like liquids. The proposed methods are versatile and can be adapted to other flocking systems, such as flamingoes, penguins, etc. |
Friday, March 6, 2020 2:03PM - 2:15PM |
X22.00015: The decision-making and mechanics of honey bee swarm formation Gary Nave, Orit Peleg In late spring and early summer, honey bee colonies may begin the process of swarming, in which the colony will send two-thirds of its population, including the mother queen, to form a new hive. The bees assemble themselves into a temporary hanging cluster for anywhere from a few hours to several days, protecting the queen and young bees while scouts identify a new nest location. While the house-hunting process has been studied, the process of formation has remained overlooked. |
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