Controllable water surface to underwater transition through electrowetting in a hybrid terrestrial-aquatic microrobot

Abstract

Several animal species demonstrate remarkable locomotive capabilities on land, on water, and under water. A hybrid terrestrial-aquatic robot with similar capabilities requires multimodal locomotive strategies that reconcile the constraints imposed by the different environments. Here we report the development of a 1.6 g quadrupedal microrobot that can walk on land, swim on water, and transition between the two. This robot utilizes a combination of surface tension and buoyancy to support its weight and generates differential drag using passive flaps to swim forward and turn. Electrowetting is used to break the water surface and transition into water by reducing the contact angle, and subsequently inducing spontaneous wetting. Finally, several design modifications help the robot overcome surface tension and climb a modest incline to transition back onto land. Our results show that microrobots can demonstrate unique locomotive capabilities by leveraging their small size, mesoscale fabrication methods, and surface effects.

Fig. 1

Design of a hybrid terrestrial-aquatic microrobot and its electrowetting pads. a A quadrupedal, 1.6 g, 4 cm × 2 cm × 2 cm hybrid terrestrial-aquatic microrbot. The robot is powered by eight piezoelectric actuators and each leg has two independent degrees-of-freedom. Each robot leg consists of an electrowetting pad (EWP) and two passive flaps. b Two passive flaps are connected to the central rigid support via compliant polyimide flexures. These passive flaps retract under drag forces opposing the robot’s heading but remain open under thrust forces in the same direction as the robot’s heading. c Perspective and front views of an EWP on the water surface. The EWP supports the robot weight via surface tension effects and the flaps paddle underwater to generate thrust forces. Scale bars (b–c), 5 mm

Fig. 2

Robot demonstration. a An illustration of robot locomotion. The robot can walk on level ground, swim on the water surface, dive into water, walk underwater, and make transitions between ground, the water surface, and the underwater environment. b Top view composite image of the robot demonstrating hybrid locomotion described in a. Scale bar, 5 cm. c Side view of the robot walking down an incline and transitioning from land to the water surface. d The robot swims on the water surface. e The robot climbs an incline when it is fully submerged in water. f The robot gradually emerges from the air–water interface. g The robot completely exits water. Scale bars (c–g), 1 cm. In c–g, two drops of blue food coloring are added to deionized water to enhance the color of water in side view images

Fig. 3

Electrowetting pad and controllable transition through the air-water interface. a Fabrication of an EWP. An EWP is laser machined from a 5 µm copper sheet, folded manually, wired, and then coated with 15 µm Parylene. b Modification of contact angle through electrowetting. When a 600 V signal is sent to the EWP, the contact angle between the EWP’s vertical sides and the water surface decreases, which reduces the surface tension force. c Spontaneous wetting of the EWP’s charged horizontal surface. The increase of surface wettability causes water to flow onto the EWP’s upper surface, consequently sinking the robot. d Composite image of a robot sinking into water when all four EWPs are actuated with a 600 V signal. Scale bars (a–d), 5 mm. e Experimental characterization of the maximum upward force generated by an EWP at different voltages. Due to change of contact angle and spontaneous wetting, the net upward force decreases as the input voltage increases

Fig. 4

Aquatic flapping kinematics and dynamics. a Swimming behavior of a diving beetle. The power stroke and the recovery stroke are asymmetric (figure taken from). b Bioinspired robot swimming kinematics feature asymmetric upstroke and downstroke without active control of the flap rotation. c Periodic control signal of the robot swing actuator is asymmetric. d Images of a single leg’s swinging motion and the passive flap rotation in water. The images are taken 0.1 period apart, corresponding to the time scale of c. Asymmetric leg swinging motion leads to favorable passive flap rotation that increases net thrust. Scale bar, 5 mm. e Comparison of experimentally measured and simulated flapping motion ψ and passive flap rotation α. f Simulated instantaneous thrust force as a function of time. The experiments and simulations shown in c–f use the same control signal. e, f show that the quasi-steady model qualitatively agrees with the experimental result, and it predicts that an asymmetric driving signal generates larger net thrust force

Fig. 5

Robot swimming and turning on the water surface. a The robot moves on the water surface at 2.8 cm s^-1 with a 5 Hz swimming gait. b The robot’s instantaneous swimming speed tracked using a high-speed video (Supplementary Movie 3). c The robot makes a complete left turn on the water surface in 13 seconds. d The robot makes a complete right turn on the water surface in 11 seconds. Scale bars (a, c, d), 2 cm. These demonstrations show that the robot can controllably move on the surface of water

Fig. 6

Robot water to land transition. a The robot is stuck at the air-water interface as it climbs an incline from underwater. The surface tension force exerts a counter-clockwise torque on the robot body, preventing the robot hind leg from lifting off. b The trajectories of the robot’s front right and hind right legs. The ramp incline is subtracted from the trajectories to show leg lift at each step. When the robot is stuck, its hind legs cannot lift off the incline surface. c The net force on a robot as it is pulled out of water vertically. The net surface tension force exceeds the robot weight. d Top and side view images of a robot getting stuck at the air-water interface. The robot hind legs splay outward due to larger surface tension force on the rear of the robot. e After stiffening robot transmission in the lift DOF, the robot moves through the air–water interface on a 3° incline. f Robot leg trajectories during the water to land transition. The ramp incline is subtracted from the trajectories. During the transition process, the lift motion of both robot front and hind legs are reduced due to the inhibiting surface tension force. The robot leg motion recovers after the robot exits the surface of water. Scale bars (a, d, e), 1 cm