Spinning-enabled wireless amphibious origami millirobot

The study addresses the lack of a wireless millimeter-scale origami robot capable of both on-ground and in-water locomotion while integrating functional tasks. Prior origami robots often require separate components for locomotion and function, increasing system complexity, and they exhibit limited locomotion modes with no amphibious capability. Rotation-based locomotion (rolling, flipping, spinning) is efficient and fast but typically requires different structural designs for ground versus water. The authors propose a magnetically actuated amphibious origami millirobot based on Kresling origami that leverages its geometry for omnidirectional rolling/flipping on land and spinning-induced propulsion in water, while using folding/unfolding as a pump for controlled liquid delivery and spinning-induced suction for cargo transport, targeting minimally invasive biomedical applications in complex hybrid environments.

Background highlights include: small-scale origami robots have shown promise for diagnosis, targeted drug delivery, and minimally invasive surgery, but usually need separate geometry for locomotion and functions and provide limited modes of motion. Rotation-based locomotion (rolling of symmetric bodies, propeller-based spinning) offers high speed/efficiency and smoother omnidirectional steering compared to crawling/wriggling/walking, yet typically demands different designs for terrestrial versus aquatic motion. No prior millimeter-scale origami device achieved both on-ground and in-water locomotion. These gaps motivate a unified, remotely actuated amphibious system leveraging rotational modes.

Design: The robot is a Kresling origami (triangulated hollow cylinder) with high global symmetry (aspect ratio H/D ≈ 1; example diameter 7.8 mm) enabling on-ground rolling and flipping, and locally tilted triangular panels acting analogously to propeller blades for spinning-based propulsion in water. A thin magnetic plate with in-plane magnetization is attached to one hexagonal end (for locomotion). For pumping, two magnetic plates with distinct in-plane magnetizations M1 and M2 (90° apart) are attached to opposite ends.

Magnetic actuation: A uniform 3D rotating magnetic field B generated by customized 3D Helmholtz coils drives rigid-body rotation via torque T = V (M × B) on the magnetic plate(s). Continuous rotation of B yields robot rotation for rolling (field rotates in a plane perpendicular to longitudinal axis), flipping (field rotates in a plane parallel to the longitudinal axis), or spinning (about the longitudinal axis in liquid). Steering is achieved by changing the rotation axis of the rotating magnetic field, reorienting the robot’s magnetization to adjust motion direction without extra mechanisms.

Self-adaptive locomotion: On unstructured terrains, the robot autonomously switches between rolling and flipping under a fixed rotating field (e.g., B = 10 mT, f = 4 Hz) to overcome obstacles such as walls, spaced stairs, and inclines, maintaining the commanded travel direction independent of its body orientation.

Pumping mechanism: The Kresling is designed monostable, recovering to the unfolded state after field removal. Applying an instantaneous magnetic field along the net magnetization direction Mnet (vector sum of M1 and M2) induces equal and opposite torques (+T, −T) on the end plates, folding the shell, contracting its internal cavity, and expelling contained liquid through designed openings. Reversing (field off) unfolds the shell, refilling the cavity. Cyclic actuation enables controlled dosing. Demonstrations used a 200 mT field to puncture an internal dye container with a needle and pump the liquid through radial cuts.

Swimming and geometry modification: For enhanced spinning propulsion, a modified Kresling includes a frontal hole and six radial cuts serving as a major inlet and outlets for flow, respectively. Right-handed tilted panels follow the right-hand rule to generate forward thrust when spinning under a rotating field (e.g., B = 10 mT at various f). Surface treatment differentiates outer hydrophobic and inner hydrophilic surfaces to facilitate water ingress and air-bubble control at interfaces.

Jumping: An instant magnetic field of magnitude B applied at angle θ relative to the plate magnetization generates an impulse enabling jumps; performance characterized at B = 40 mT and varying θ (e.g., θ = 120° demonstrated).

Cargo capture: Spinning generates low frontal pressure at the hole, producing suction to capture and transport solid cargo; release occurs by orienting the hole downward.

Experimental environments: Locomotion trials included flat and unstructured terrains, hybrid terrestrial–aquatic testbeds, and ex vivo pig stomachs (with rugae, mucosa, and viscous fluid of 12 mPa·s). Magnetic field examples: on-ground locomotion B = 10 mT, f = 2–4 Hz; underwater swimming B ≈ 10 mT, f up to 30 Hz; viscous-fluid swimming B ≈ 12 mT, f = 24 Hz.

CFD simulations: Using Ansys Fluent, rigid-body models with and without the frontal hole/cuts were analyzed under imposed rotational speed and external flow corresponding to experimental spinning/swimming. Streamlines and normalized pressure fields were compared to assess frontal pressure and drag differences.

Fabrication: Kresling patterns cut from 0.05 mm polypropylene; inner surface hydrophilic coated; Mylar hexagons (0.127 mm) attached, one with a 3-mm frontal hole, plus six radial cuts. Magnetic plates: Ecoflex-0030 silicone embedded with 10 vol% NdFeB particles (avg 100 µm) and 20 vol% glass bubbles. A cylinder N52 permanent magnet (50 mm diameter × 25 mm thickness) was used for high-field folding in pumping demonstrations.

• Multimodal, amphibious locomotion: A single origami body achieves on-ground rolling and flipping and in-water spinning-induced propulsion. The robot autonomously switches between rolling and flipping to traverse unstructured terrains (walls, spaced stairs, inclines) under a fixed rotating field (e.g., B = 10 mT, f = 4 Hz), maintaining the commanded direction regardless of body orientation.
• Jumping: Instant-field actuation (B = 40 mT) with θ = 120° enabled jumping; performance as a function of θ characterized, with demonstrated height and distance (figure indicates heights up to ≈75 mm under tested conditions).
• Controlled pumping/drug delivery: Cyclic folding/unfolding using two end magnets (M1 ⟂ M2, 90°) yields a reversible pump; a 200 mT field contracted the shell, punctured an internal dye container with a needle, and expelled liquid through radial cuts for controllable dosing. Ex vivo pig stomach experiments demonstrated locomotion to a target and controlled release at location.
• Underwater swimming performance: Modified Kresling with frontal hole and radial cuts achieved higher speed than without modifications. Maximum measured speeds: 81.2 mm/s (11.9 body lengths/s) at B = 10 mT, f = 30 Hz with hole/cuts vs 66.0 mm/s (9.7 body lengths/s) without. Speed increased with rotation frequency.
• CFD insights: The frontal hole eliminates the stagnation point and lowers frontal pressure, enabling fluid capture and centrifugation through radial cuts, reducing swimming resistance and increasing speed versus the solid-front design.
• Navigation agility: Magnetic steering enabled precise 1D, 2D (∞ path), and 3D spiral paths underwater; transitions across ground–water–interface environments were demonstrated. Interface swimming and floating achieved by self-capturing an air bubble at the frontal hole to adjust effective density; reorientation allows bubble release and submergence.
• Cargo transport: Spinning-induced low-pressure suction at the frontal hole enabled capture, transport, and release of solid cargo in hybrid terrestrial–aquatic scenarios.
• Ex vivo demonstrations: In an ex vivo pig stomach with viscous fluid (12 mPa·s), the robot performed self-adaptive rolling/flipping on rugae (B = 10 mT, f = 2 Hz) and swimming with steering (B ≈ 12 mT, f = 24 Hz).

The Kresling-based, magnetically actuated millirobot leverages geometrical symmetry and tilted panels to unify rotation-enabled locomotion on land (rolling, flipping) and in water (spinning propulsion) within a single, simple body. The foldable thin-shell architecture provides an internal cavity for functional integration, enabling controlled liquid delivery via reversible folding/unfolding and targeted cargo transport via spinning-induced suction. Magnetic actuation affords omnidirectional steering without additional mechanisms and supports autonomous terrain adaptation by self-switching locomotion modes. Underwater modifications (frontal hole, radial cuts) reduce frontal pressure and resistance, enhancing swimming speed. Ex vivo experiments in stomach tissue and viscous media indicate feasibility for minimally invasive operations in complex biomedical environments. The platform is amenable to scaling and functional integration (e.g., onboard tools) for broader clinical tasks.

A wireless amphibious origami millirobot was demonstrated that integrates multimodal rotation-enabled locomotion (rolling, flipping, spinning) with functional capabilities for controlled liquid delivery and cargo transport, operated by external magnetic fields. The design exploits Kresling origami geometry and folding mechanics to unify locomotion and function in one body and shows robust performance across terrestrial, aquatic, and interface environments, including ex vivo demonstrations. Future directions include investigating fluid viscosity effects on swimming performance, scaling the concept down for intraluminal biomedical environments (e.g., blood vessels, ureters), and integrating onboard components (e.g., cameras, forceps) to enable tasks such as endoscopy and biopsy.

The study primarily demonstrates feasibility with benchtop and ex vivo experiments; in vivo validation is not reported. The influence of fluid viscosity on swimming performance is acknowledged but left for future in-depth study. Performance metrics (e.g., jumping, dosing rates) are characterized under specific magnetic field strengths and configurations, which may require adaptation for different clinical environments.