Session W22: Robophysics III

Collective Robot Sex on Dynamic Resource Landscapes

We present a community of robots (``Jeeps''), which move over a resourc landscape consisting of a large light-emitting diode (LED) light board whose local RGB intensity represents fitness, the local intensity can change due to robot local presence, making the landscape dynamic. Each Jeep has a basic digital genotype that carries the response (phenotype) to a given local fitness value, down-ward firing RGB sensors to measure local intensity and RGB color, and side-firing sensors and LEDs to communicate with neighboring Jeeps and engage in gene exchange (robot sex) with neighboring Jeeps. The Jeeps move in response to the current local fitness gradient given by the gradient in color intensity of their position on the light board. The jeeps ``genes’’ mutate at a rate inversely proportional to the color intensity at their position, breed genomes, and consume resources. This robot community has great generality which spans the space of many-body soft-matter physics, evolutionary biology and multi-cell disease states. We show robots on the dynamic landscape exhibit the self-organization of a collective, distributed intelligent complexity   

Delay Induced Swarm Pattern Bifurcations in Mixed Reality Experiments

Swarms of coupled mobile agents with communication-time delay are known to exhibit multiple dynamic patterns in space, which depend on interaction strength and delay. We demonstrate experimentally delay-induced bifurcations between spatio-temporal patterns with two separate robotic platforms using a mixed-reality framework, and isolate parameter regions where transitions occur. We show that multiple rotation patterns persist in the presence of collision-avoidance and uncertainties in the real robot dynamics. Moreover, we show the existence of multi-stability of rotational patterns not predicted by usual mean field dynamics. Our experimental results increase understanding of swarm-pattern robustness in the physical world, and may form the basis for improved reproducible models of swarms with physical robots.   

Robotic control of live jellyfish swimming to enhance propulsion

Animal locomotion and bioinspiration have the potential to expand the performance capabilities of robots, but current implementations are limited. Mechanical soft robots leverage engineered materials and are highly controllable, but these biomimetic robots consume more power than corresponding animal counterparts. Biological soft robots from a bottom-up approach offer advantages such as speed and controllability, but are limited to survival in cell media. Instead, biohybrid robots that comprise live animals and self-contained microelectronic systems leverage the animals’ own metabolism to reduce power constraints and body as an natural scaffold with damage tolerance. We demonstrate that by integrating onboard microelectronics into live jellyfish, we can enhance propulsion up to threefold, using only 10 mW of external power input to the microelectronics and at only a twofold increase in cost of transport to the animal. This robotic system uses 10 to 1000 times less external power per mass than existing swimming robots in literature, and can be used in future applications for ocean monitoring.   

Neuromuscular actuation of biohybrid motile bots

The integration of muscle cells with soft robotics in recent years has led to the development of biohybrid machines capable of untethered locomotion. A major frontier that currently remains unexplored is neuronal actuation and control of such muscle-powered biohybrid machines. As a step toward this goal, we computationally designed, optimized, and implemented light-sensitive flagellar swimmers driven by on-board neuromuscular units. The body of the swimmer consists of a free-standing soft scaffold, skeletal muscle tissue, and optogenetic stem cell-derived neural cluster containing motor neurons. Neural stimulation triggers cyclic muscle contractions, driving time-irreversible flagellar dynamics, thereby providing thrust for untethered forward locomotion of the swimmer. Overall, this work demonstrates an example of a biohybrid robot implementing neuromuscular actuation and illustrates a path toward the forward design and control of neuron-enabled biohybrid machines.   

Gait coordination and hydrodynamic performance of a quadriflagellate robophysical model

Multi-legged animals coordinate their limbs in distinctive patterns known as gaits. At the microscopic scale, quadriflagellate algae have been found to exhibit similar capabilities, coordinating their flagella to numerous rhythmic patterns to generate propulsion (Wan & Goldstein, 2016). To study quadriflagellate gait coordination, we developed a robophysical model which replicates quadriflagellate swimming at low-Reynolds number. We focus on two distinct gaits, the pronk and the trot, and explored the effects of flagellar orientation. When the flagella were oriented parallel to the cell body, forward motion was measured at 0.30±0.09 body lengths per gait cycle (BL/cyc) for the trot and at 0.19±0.03 BL/cyc for the pronk. Results are comparable to microorganisms’ performance, where using the trot enables a higher speed (0.39±0.18 BL/cyc) than the pronk (0.18±0.05 BL/cyc). Surprisingly, when the flagella beat planes were perpendicular to the cell body, hydrodynamic performance improved significantly. Forward motion was measured at 0.66±0.13 BL/cyc for the trot and at 0.38±0.10 BL/cyc for the pronk. The results show that hydrodynamic performance is highly sensitive to swimming gait and flagellar orientation.   

Locomotion of Soft Robots with Flexible Uni-flagellum in low Reynolds Number Fluid

In nature, more than 90% of the bacteria propel using a single flagellum. The uniflagellar bacterium is comprised of a basal body and a flagellum, with a flexible hook connecting the two. Experimental and numerical observations indicate that the flexibility of the flagellum and its buckling are essential to change the swimming direction of the bacterium. We simulate the locomotion and deformation of the flagellated system in low Reynolds number conditions using a combination of the Discrete Elastic Rods and the Lighthill Slender Body Theory. The simulator shows that the robot is able to follow a prescribed three-dimensional trajectory by simply controlling the angular velocity of the flagellum. Inspired by natural bacterial structure, we fabricate a self-contained robot with a rigid and a motor-actuated elastomeric helical rod - our analog for the flagellum. With a motor, sensors, and a flexible control circuit embedded inside the head, the robot is able to follow a prescribed nonlinear trajectory by flagellum buckling. This robotic system demonstrates how structural instability can be harnessed for control of soft robots. It can also serve as an analog model for the natural bacterium to help us understand the fluid dynamics and biophysics underlying the propulsion of bacteria.   

Simply controlled 1DOF adaptive soft robotic fish with multiple swimming behaviors

Research in robotic grasping and soft machines has demonstrated that the use of “smart” mechanical structures in robotic systems can enable complex behaviors in dynamically changing, uncertain environments without the need for complicated, low-level control strategies. Upon that premise, we designed an 1DOF simply controlled robotic fish to explore the ability to achieve different locomotion patterns and passively generate adaptive swimming behaviors, simply by changing the servo’s output parameters. This 1-DOF, cable-driven robotic fish mimics the cascaded skeletal structure of soft spine tissue and hard spine bone seen in many fish species using a series of rigid segments and elastic joints. By changing the frequency of actuator motion, the robotic fish can not only change its lateral swimming pattern from oscillatory (2.3cm/s) to undulatory (11.0cm/s), but also change swimming direction from forward to backward and steering by biasing servo sweeping offset. Preliminary tests also show that the robot can swim downward to 7m successfully. In addition, the soft body helps the fish passively navigate through gaps and around obstacles without any feedback control, greatly enhancing the locomotion robustness in complex, cluttered environments.   

Magnetic Actuation and Biological Sensing for Soft Micro Bio Robots

Small-scale robotic systems have the potential to drive technological development in biological research and medicine. Similar in size to insects or as small as a single cell, robots at these scales have manipulated cells and tissues, delivered therapeutics, and monitored biological environments. In order to perform these tasks, small-scale robots must be able to precisely locomote, take measurements, and use those measurements to make decisions. In this talk, we will discuss our distinct approach to actuation and sensing through the design and fabrication of soft micro bio robots composed of an organic hydrogel embedded with iron oxide nanoparticles. Due to its biocompatible, soft, and porous scaffold, this robot can transport and deliver chemical, cellular, and bimolecular payloads. We drive these robots using uniform, rotating magnetic fields across a range of rotational frequencies and within various fluids spanning three orders of magnitude in viscosity. Inspired by developments in synthetic biology, we genetically engineer sensors and processing units harbored within bacteria which are grown and function on-board the robot. These results showcase a hybrid strategy to actuate and integrate sensors on-board biologically relevant small-scale robots.   

Surface electrochemical actuators for micron-scale fluid pumping and autonomous swimming

Recently, our group demonstrated a new class of electronic actuators called surface electrochemical actuators (SEAs). They use surface adsorption on a nanometer-thick cantilever to produce micron-scale radii of curvature with only fractions of a volt for actuation. Here we use SEAs to mechanically pump fluid at the micron-scale in several different geometries, both in the rigid panel (Purcell) and flexible (flagella) limits. We further discuss ongoing work to integrate SEAs with photovoltaics to create optically powered swimmers. Our ultimate goal is to create fully autonomous swimming microrobots with onboard electronics that can sense and respond to their environment in complex ways, yet are too small to be resolved by the naked eye.   

Material remodeling on granular terrain yields robustness benefits for a robophysical rover

Planetary rovers face the risk of entrapment during extraterrestrial exploration; these risks led NASA JSC to develop RP-15, a 300 kg rover capable of lifting and sweeping each wheel to execute a crawling behavior to escape entrapment. We created a miniature rover (2.1 kg) as a robophysical model of RP-15 and conducted systematic and automated experiments in a tilting bed containing a model granular medium (poppy seeds) to discover gaits which allow progress in various conditions. A combination of stepping, paddling, and wheeling (“RS gait”) allows climbing of loose granular slopes. The RS gait produces up to 4x greater tractive (drawbar) forces than wheel rotation alone on dry granular media. Drawbar results were validated through experiments on RP-15 at JSC. On slopes near the maximum angle of stability, a different gait used the front wheels as agitators while paddling the rear wheels to climb granular slopes successfully via controlled avalanching. While substrate disturbance typically hinders granular locomotion, using appropriate active granular reconfiguration can create localized granular structures that facilitate effective locomotion.   

Investigating Growth and Granular Fluidization in a Minimally Invasive Burrowing Robot

Subterranean navigation is simply hard to do. The forces resisting underground motion are orders of magnitude higher than in air or water. Here we present a paradigm for minimally invasive robotic burrowing that results in movement an order of magnitude faster and deeper than previous low-impact approaches. Three principles enable this behavior: tip-extension to eliminate skin drag; granular fluidization to reduce form drag; and granular fluidization to control lift forces, a heretofore unstudied phenomenon. We present experimental results from controlled intrusion tests, studying the effects of fluidization direction, depth, and flow rate on both drag and lift. We show that lift is highly dependent on fluidization angle (i.e., the direction of air flow at the tip), decreasing as the fluidization is changed from horizontal to vertically downward, but drag reduction is insensitive to fluidization angle. With the use of pneumatic artificial muscles for steering, we demonstrate a functional burrowing robot, capable of navigating through sand and subterranean obstacles. Our results advance the understanding and capabilities of robotic subterranean locomotion, and the forces at play.   

Force response of running up a sand dune

Running up a sand dune is challenging because: (1) sand fluidizes when an external force exceeding the material yield stress is applied; and (2) at inclines approaching the angle of repose, the sand pile is increasingly unstable. In this study, we experimentally examined the impact force normal to a flat plate against a bed of poppy seeds, to determine how intrusion kinematics affect force generation. We tested a range of impact speeds (0.01-1.2 m/s), substrate angles, and impact angles. We identified two regimes with distinct force production patterns: 1) the gravity regime (<0.6 m/s) and 2) the inertial regime (>=0.6 m/s). In the gravity regime, the force-depth relation tends to diverge when comparing, for example, (θ_intrusion, θ_substrate)=(40,0) and (0,40). This difference is likely due to the propensity of the angled substrate to avalanche and reduce the material force response. On the other hand, the same set would converge at higher impact speeds, indicating that inertia dominates the system with particle movement instigated by the intruder exceeding that due to avalanching. This insight will help inform how animals and robots may navigate up sand dunes efficiently.   

Generation GrowBots: learning from plants how to design self-morphing, growing robots

Plants show unique capabilities of endurance and the ability to adapt their morphology to different environmental conditions. This plasticity is materialized through a variety of strategies, including moving-by-growing to search for nutrients, lights, or for external supports; rapid movements to capture prays; or passive movements for spreading seeds and fruits. Together with plant biologists and materials scientists, we are deeply investigating the biomechanics, materials, energy efficiency mechanisms, and behavior of a variety of plant species. For the first time, we have proposed a growing robot inspired by the movements and the behaviors of plant roots, able to create its own structure exploiting a 3D printer-like system integrated into its tip and depositing a thermoplastic material. With a focus on climbing plants, we are now taking inspiration from their material properties, biomechanics, and searching&anchoring capabilities for the design of new multi-functional, self-morphing, adaptable, growing robots. This new generation of plant-inspired “growbots”, able to self-create their structure, will find potential applications in a variety of sectors, including the exploration and monitoring of archeological sites, future urban architecture, or extra-terrestrial areas.