An official website of the United States government
Soft robots, known for their compliance and deformable nature, have emerged as a transformative field, giving rise to various prototypes and locomotion capabilities. Despite continued research efforts that have shown significant promise, the quest for energy-efficient mobility in soft-limbed robots remains relatively elusive. We introduce a discrete locomotion gait called “tumbling,” designed to conserve energy and implemented in a topologically symmetric soft-limbed robot. The incorporation of tumbling enhances the overall locomotive abilities of soft-limbed robots, offering advantages such as increased agility, adaptability, and the ability to correct orientation, which are essential for navigating non-engineered environments that include natural-like irregular terrains with obstacles. The principle behind tumbling locomotion involves a deliberate shift in the robot's center of gravity in the direction of motion, guided by the kinematics of its soft limbs. To validate this locomotion strategy, we developed a robot simulation model operating within a virtual environment that incorporates physics and contact interactions. After optimizing the tumbling locomotion strategy through simulations, we conducted experimental tests on a physical robot prototype. The experiments validate the effectiveness of the proposed tumbling gait. We conducted an energy cost analysis to compare the tumbling locomotion with the previously reported crawling gait of the robot. The results of this analysis demonstrate that tumbling represents an energy-efficient mode of locomotion for soft robots, saving up to 60% and 65% energy than crawling locomotion on flat and natural-like irregular terrains, respectively.
more »
« less
This content will become publicly available on June 30, 2026
Body-Induced Soft Robotic Locomotion: a Sim-To-Real Approach
Soft robots, valued for their compliance and deformable nature, have demonstrated their outstanding abilities in complex environments. However, the nonlinear dynamics make it challenging to derive efficient locomotion patterns from analytical methods. This is largely due to the high computational cost associated with simulating soft-bodied models. Conversely, rigid-body models, such as those used in Gazebo, offer computational efficiency but cannot directly represent soft robots. We address these challenges by introducing customized Gazebo plugins that enable the simulation and analysis of soft robot locomotion dynamics. These plugins are complemented by a novel JointStiffnessPlugin, integrated with ROS services, for fine-tuning effort-controlled parameters. The system identification process is followed to match the simulation dynamics with the real soft robot to minimize the sim-to-real gap. Utilizing the proposed simulation framework and Bayesian Optimization, we derived a body-induced locomotion strategy that achieves enhanced efficiency. This strategy, relying solely on periodic spine bending and robot pose for forward propulsion, demonstrably reduces energy consumption compared to conventional gaits. Experimental results confirm a 42 % energy expenditure reduction relative to four-legged crawling.
more »
« less
- PAR ID:
- 10633673
- Publisher / Repository:
- IEEE
- Date Published:
- ISBN:
- 979-8-3315-2441-8
- Page Range / eLocation ID:
- 535 to 541
- Format(s):
- Medium: X
- Location:
- College Station, TX, USA
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Soft robots that can harvest energy from environmental resources for autonomous locomotion is highly desired; however, few are capable of adaptive navigation without human interventions. Here, we report twisting soft robots with embodied physical intelligence for adaptive, intelligent autonomous locomotion in various unstructured environments, without on-board or external controls and human interventions. The soft robots are constructed of twisted thermal-responsive liquid crystal elastomer ribbons with a straight centerline. They can harvest thermal energy from environments to roll on outdoor hard surfaces and challenging granular substrates without slip, including ascending loose sandy slopes, crossing sand ripples, escaping from burying sand, and crossing rocks with additional camouflaging features. The twisting body provides anchoring functionality by burrowing into loose sand. When encountering obstacles, they can either self-turn or self-snap for obstacle negotiation and avoidance. Theoretical models and finite element simulation reveal that such physical intelligence is achieved by spontaneously snapping-through its soft body upon active and adaptive soft body-obstacle interactions. Utilizing this strategy, they can intelligently escape from confined spaces and maze-like obstacle courses without any human intervention. This work presents a de novo design of embodied physical intelligence by harnessing the twisting geometry and snap-through instability for adaptive soft robot-environment interactions.more » « less
-
Tensegrity robots, composed of rigid rods and flexible cables, exhibit high strength-to-weight ratios and significant deformations, which enable them to navigate unstructured terrains and survive harsh impacts. They are hard to control, however, due to high dimensionality, complex dynamics, and a coupled architecture. Physics-based simulation is a promising avenue for developing locomotion policies that can be transferred to real robots. Nevertheless, modeling tensegrity robots is a complex task due to a substantial sim2real gap. To address this issue, this paper describes a Real2Sim2Real (R2S2R) strategy for tensegrity robots. This strategy is based on a differentiable physics engine that can be trained given limited data from a real robot. These data include offline measurements of physical properties, such as mass and geometry for various robot components, and the observation of a trajectory using a random control policy. With the data from the real robot, the engine can be iteratively refined and used to discover locomotion policies that are directly transferable to the real robot. Beyond the R2S2R pipeline, key contributions of this work include computing non-zero gradients at contact points, a loss function for matching tensegrity locomotion gaits, and a trajectory segmentation technique that avoids conflicts in gradient evaluation during training. Multiple iterations of the R2S2R process are demonstrated and evaluated on a real 3-bar tensegrity robot.more » « less
-
Tensegrity robots, composed of rigid rods and flexible cables, exhibit high strength-to-weight ratios and significant deformations, which enable them to navigate unstructured terrains and survive harsh impacts. They are hard to control, however, due to high dimensionality, complex dynamics, and a coupled architecture. Physics-based simulation is a promising avenue for developing locomotion policies that can be transferred to real robots. Nevertheless, modeling tensegrity robots is a complex task due to a substantial sim2real gap. To address this issue, this paper describes a Real2Sim2Real (R2S2R) strategy for tensegrity robots. This strategy is based on a differentiable physics engine that can be trained given limited data from a real robot. These data include offline measurements of physical properties, such as mass and geometry for various robot components, and the observation of a trajectory using a random control policy. With the data from the real robot, the engine can be iteratively refined and used to discover locomotion policies that are directly transferable to the real robot. Beyond the R2S2R pipeline, key contributions of this work include computing non-zero gradients at contact points, a loss function for matching tensegrity locomotion gaits, and a trajectory segmentation technique that avoids conflicts in gradient evaluation during training. Multiple iterations of the R2S2R process are demonstrated and evaluated on a real 3-bar tensegrity robot.more » « less
-
Soft robots hold significant potential in legged locomotion due to their inherent deformability, enabling enhanced adaptability to various environmental conditions and the generation of diverse locomotion gaits. While various soft robots have been proposed for terrestrial locomotion, research on dynamically-stable locomotion, such as trotting, with actuated soft bending limbs remains limited. We introduce a pneumatically-actuated soft quadruped featuring a soft body capable of a variety of dynamically-stable trotting locomotion. We utilize soft limb kinematics and parameterize fundamental limb locomotion to obtain quadrupedal locomotion trajectories for both linear and curvilinear motions. We also employ a physics-enabled dynamic model to optimize and evaluate trotting locomotion trajectories for dynamic stability. We further validate the stable locomotion trajectories through empirical experiments conducted on a soft quadruped prototype. The results demonstrate that the quadruped trots at a peak speed of 1.24 body lengths per second when traversing flat and uneven terrains, including slopes, cluttered areas, and naturalistic irregular surfaces. Furthermore, we compare the energy efficiency between trotting and crawling locomotion. The findings reveal that trotting is significantly more energy-efficient than crawling, with an average energy saving of up to 42%.Note to Practitioners—This paper was motivated by the challenge of achieving dynamically stable and efficient locomotion in soft quadrupeds. Many soft-legged robots are typically designed for statically stable, albeit inefficient and slow, locomotion gaits such as crawling. Our research aims to address this practical challenge of improving mobility in soft-legged robots. We develop a novel soft quadruped with pneumatically-actuated soft limbs that achieves efficient trotting that is 42% more energy-efficient than crawling. This work is particularly relevant for industries requiring adaptable and efficient navigation in environments, such as search and rescue, agricultural monitoring, and exploration. The development and optimization of trotting gaits through a physics-enabled dynamic model for dynamic stability provide a foundational framework for enhancing the adaptability and operational utility of soft robots. While our findings mark a significant step forward, challenges remain in deploying these locomotion strategies on autonomous untethered robots with onboard sensor feedback. Future research will focus on these areas, aiming to improve the practical deployment and robustness of soft robotic locomotive systems.more » « less
