Muscle-inspired elasto-electromagnetic mechanism in autonomous insect robots

Abstract

In nature, the dynamic contraction and relaxation of muscle in animals provide the essential force and deformation necessary for diverse locomotion, enabling them to navigate and overcome environmental challenges. However, most autonomous robotic systems still rely on conventional rigid motors, lacking the adaptability and resilience of muscle-like actuators. Existing artificial muscles, while promising for soft actuation, often require demanding operational conditions that hinder their use in onboard-powered small autonomous systems. In this work, we present the Elasto-Electromagnetic mechanism, an electromagnetic actuation strategy tailored for soft robotics. By structuring simple elastomeric materials, this mechanism mimics key features of biological muscle contraction and optimizes actuation properties. It achieves significant output force (~210 N/kg), large contraction ratio (up to 60%), rapid response (60 Hz), and low-voltage operation (<4 volts) within a robust, miniaturized framework. It also enhances energy efficiency by maintaining stable states without continuous power input, similar to catch muscles in mollusks. The resulting insect-scale soft robots, therefore, demonstrate adaptive crawling, swimming, and jumping, autonomously navigating open-field environments. This muscle-inspired electromagnetic mechanism, facilitated by elastic structural variations, expands the autonomy and functional capabilities of small-scale soft robots, with potential applications in rescue and critical signal detection.

Fig. 1. Muscle-inspired elasto-electromagnetic (EEM) mechanism for insect-scale soft robots.

a The maximum output force and contraction ratio of animals and autonomous actuators at insect-scale. The contraction ratio of animal muscles is between 10% and 30%. b Inspired by the contraction and relaxation of muscles, the EEM mechanism mimics these dynamics by balancing the magnetic attractive force, $F_{\mathrm{e}\mathrm{m}}$, and the elastic response of the soft structure, $F_{\mathrm{e}}$. c Inspired by the structure and locomotive gaits of small animals, autonomous soft robots are designed as an inchworm crawler, legged crawler, and swimmer, each showcasing high force output, large contraction ratios, and enhanced energy efficiency. These autonomous small robots with sensors integrated show great potential in applications like search-and-rescue, cave discovery, and environmental detection. Source data for (a) is provided as a Source Data file.

Fig. 2. The muscle-inspired EEM actuation mechanism.

a Schematic illustration of the EEM actuation system. (i) The system is composed of a magnetized hard magnet, an elastomeric structure, and an electrical coil intertwined with soft magnets. The variable $x$ represents the actuation displacement and $x_{\mathrm{limit}}$ is the physical constrained position. (ii) The EEM system relies on a delicate balance between the magnetic attractive force, $F_{\mathrm{e}\mathrm{m}}$ and the elastic response, $F_{\mathrm{e}}$, with their force-displacement curves presented in the plot. The intersection represents the initial balance position (x₀), with the system capable of reaching a displacement limit ($x_{\mathrm{limit}}$) at maximum force exertion. b Force-balance modulation by adjusting the electric current for actuation. Upon applying a current signal in the coil, the force–displacement curve of $F_{\mathrm{em}}$ alters upwards, resulting in the system to reach a new balance position. c Force-balance modulation via designing the elastomeric structure for customizing different actuation properties. (i)–(iii) The EEM actuation system can be customized to have one, two, or even three stable states within a compact design. The elastomeric structures and force-displacement relations are presented, with the purple dot indicating their stable states. (iv) and (v) Similarly, the EEM system can be tailored with either a smooth transition property or a step transition property. The system with a step transition can produce a large energy release for enhancing dynamic performance, such as enabling a jumping motion. (vi) The contraction ratio of the EEM system can be designed to reach 60%. Source data for c are provided as a Source Data file.

Fig. 3. The EEM motor and joints.

a The schematic illustration of the EEM motor. A flexure linkage mechanism is integrated in the elastomeric structure such that it can convert the actuator’s linear contraction/relaxation motion into rotational movement in the linkage. The EEM motor can be easily scaled to millimeter size and decimeter size. b The bistable EEM motor. (i) The actuation strategy and the output force definition of the EEM motor. $F_{\mathrm{e}\mathrm{m}}$ represents the magnetic attractive force and $F_{\mathrm{e}}$ represents the elastic response. (ii) The maximum and average output forces of the bistable motor during the contraction and relaxation phases across various actuation currents. c The EEM motors are modified to construct the translational joint for contraction and the rotational joint for flapping. Source data for (b) is provided as a Source Data file.

Fig. 4. The autonomous insect-scale soft inchworm crawler.

a Components of the autonomous clawer, which includes an inchworm-like crawling joint, a battery, and a control circuit. b The autonomous insect-scale robot can crawl on different surfaces (PVC plate, glass, wood, stone surfaces) and even in an open-field environment with soil ground. The pictures are superimposed frames from Supplementary Video 7. c The performance analysis of the inchworm crawler robot. (i) The large stride of the crawler robot, reaching 35%. (ii) The roughness of the crawling surfaces in the experiments. (iii) The crawling speeds across different surfaces, under actuation frequencies ranging from 1 to 5 Hz. (iv) The crawling speeds on inclined PVC surfaces with slopes of 5°, 10°, and 15° across various actuation frequencies. d Continuous operation. (i) Sustained time of the autonomous crawler under different actuation frequencies with a single battery. The yellow dots represent the experimental evaluation results, and an inverse proportional relationship is fitted to the plot by the light blue curve. T and f represent the operation time and actuation frequency, respectively. It is worth noting that at the actuation frequency of 0.67 Hz, the crawler can operate for more than one hour. (ii) The recorded temperature rise of the crawling robot under continuous operation across different frequencies. The temperature reaches thermal equilibrium over time, with a slight decrease observed later due to battery voltage drop as power is consumed. Source data for (c) and d are provided as a Source Data file.

Fig. 5. The autonomous insect-scale soft inchworm crawler and legged crawler.

a The experiment to evaluate the high impact resistance capability of this untethered soft EEM inchworm crawler. The robot first crawls on a platform at a 30 m height from the ground, and then drops to the ground with a high impact. After that, this autonomous crawler can continue to crawl forward without damage. b The inchworm robot can also change its crawling direction by integrating two independent coils into its body while not increasing the robot's size. c The autonomous soft-legged crawler. (i) Components of the robot, including a legged crawling joint, a battery, and a control circuit. (ii) The legged crawling locomotion in a lab environment and an open-field environment (soil ground). The pictures are superimposed frames from Supplementary Video 9.

Fig. 6. The autonomous insect-scale soft swimmer and jumper.

a The autonomous soft swimmer. (i) Components of the robot, including a swimming joint, a battery, a control circuit, and buoyance components for flotation. (ii) The soft robot swims in lab environment. (iii) The swimming speeds of the robot correspond to different time ratios spent on the upstroke and downstroke phases with a flapping frequency of 1 Hz. (iv) The swimming speeds of the robot correspond to different flapping frequencies with the time ratio of 9:1, 1:1, and 1:9. b The autonomous soft robot swims on an open-field river surface with a speed of 24 mm/s at the flapping frequency of 0.67 Hz. c Directional swimming capability demonstration via integrating two independent coils in the robot. d The autonomous insect-scale soft jumper. (i) Components of the robot, including a jumping joint, a battery, and a control circuit. (ii) This untethered autonomous jumper demonstrates a unique capability of jumping multiple times on this length scale. The pictures are superimposed frames from Supplementary Video 11, and each frame is shifted left for a better illustration of the jumping motion. Source data for (a) are provided as a Source Data file.

Fig. 7. The autonomous insect-scale robots with signal detection capability.

a Autonomous clawer for physical signals detection. In this experiment, we integrate a temperature and humidity sensor in the control circuit of the soft inchworm crawler and then command the robot to navigate through a narrow and complex pathway. Among the pathways, we create two constrained chambers (A and B) with one of them to have a high humidity environment and the other to have a low temperature environment. The detected humidity and temperature signals are plotted in the figures, indicating an efficient sensing capability. b Equipped with an ethanol gas sensor, this insect-scale soft robot was navigated through a narrow channel into an enclosed chamber housing an open container of liquid ethanol. Upon entering the chamber, the robot’s sensor promptly detected the ethanol vapors, successfully transmitting this chemical signal externally. Source data for (a) and b are provided as a Source Data file.