Underwater Bio-Inspired Soft Robots

• Underwater bio-inspired soft robots are engineered systems that replicate the flexibility and adaptive morphologies of marine organisms for complex aquatic tasks.
• They employ biohybrid actuation mechanisms—such as fluidic, dielectric, and origami-inspired designs—integrated with soft materials to enhance maneuverability and environmental compatibility.
• Advances in sensing, control, and rapid prototyping enable these robots to perform ecological monitoring, delicate manipulation, and eco-friendly operations in unstructured underwater settings.

Underwater bio-inspired soft robots are engineered systems that leverage the compliance, multifunctionality, and adaptive morphologies found in marine organisms to achieve enhanced performance, environmental compatibility, and novel operational capabilities in aquatic environments. These robots exploit architectural, actuation, and control paradigms inspired by fish, cephalopods, jellyfish, echinoderms, bacteria, and other aquatic life forms, and have emerged as an enabling technology for oceanexploration, ecological monitoring, environmental remediation, and complex manipulation in unstructured underwater settings (Wang et al., 18 Jan 2026, Li et al., 16 Aug 2025).

1. Fundamental Biological Inspirations and Biohybrid Principles

Marine organisms provide a diverse repertoire of locomotion, adhesion, manipulation, and sensory strategies for underwater robotics. Architectures such as muscular hydrostats (octopus arms), fin-and-body undulators (fish), soft-bell pulsers (jellyfish), jet-propelled mantles (squid), and flagella-driven bacteria are directly transposed into robotic analogues via compliant, distributed material systems.

• Muscular hydrostat models: Octopus arms exhibit unconstrained continuum bending, torsion, and telescopic extension through three orthogonally arranged muscle layers constrained at constant volume. Soft robotic manipulators replicate these through cable networks or embedded fluidic chambers in elastomeric matrices (Li et al., 16 Aug 2025, Walker et al., 2024).
• Undulatory swimming and fin propulsion: BCF (Body/Caudal Fin) and MPF (Median/Paired Fin) locomotion is emulated via elastomeric tails with antagonistic chambers (hydraulic or pneumatic), distributed smart-material "muscles" (e.g. HASEL, DEA, SMA), and sophisticated kinematic sequencing, enabling traveling or standing waveforms tailored for thrust or maneuverability (Wang et al., 18 Jan 2026, Soto et al., 15 Apr 2025, Zhang et al., 2023).
• Jet propulsion: Cephalopod-inspired robots employ elastomeric mantles or combustion-driven chambers to realize pulsed-jet thrust and escape maneuvers, combining high instantaneous force with structural compliance (He et al., 2023, Chen et al., 16 Apr 2025).
• Flagellar and ciliary mechanisms: Bacteria-analogues such as ZodiAq and BactoBot employ soft silicone helical flagella driven by distributed actuators, achieving omnidirectional, low-disturbance propulsion suited to delicate ecosystems (Chowdhury et al., 25 Sep 2025, Mathew et al., 25 Mar 2025).
• Soft adhesion: Remora- and mollusk-inspired suction actuators employ bilayer doming, switchable negative pressure, and contact-compliant lips to adhere on wet, rough, or underwater surfaces, supporting amphibious climbing and biohitchhiking analogues (Tang et al., 2018).

2. Soft Material Platforms and Fabrication Methods

Material selection and characterization are critical for achieving biological fidelity and operational robustness:

• Elastomers: Ecoflex, DragonSkin, PDMS, and Mold Star silicones (E≈0.1–3 MPa) are ubiquitous for moderate modulus, extensibility, and stability (Singh et al., 13 Nov 2025, Zhang et al., 2023).
• Hydrogels and LCEs: Employed to mimic tissue-like softness and environmental responsiveness, enabling transparency, biocompatibility, and even self-healing (Wang et al., 18 Jan 2026, Li et al., 16 Aug 2025).
• Smart materials: HASEL and DEA actuators provide high strain rates, muscle-like contraction, and electrical tunability, while SMAs and SMPs afford thermal shape changes and high blocking force (Soto et al., 15 Apr 2025, Zhang et al., 2023).
• Origami and composite laminates: The Yoshimura fish-mouth gripper exemplifies crease-based mechanical intelligence for robust underwater manipulation, fabricated from patterned composite sheets (nylon–TPU, PETG) (Guo et al., 14 Mar 2025).
• Modular casting and rapid prototyping: Lost-wax, 3D printing, and soft lithography enable fast iterations on chamber layouts, stiffness gradients, and integration of electronics or sensors (Zhang et al., 2022, Singh et al., 13 Nov 2025).

3. Actuation Mechanisms and Kinematic Control

A diverse suite of actuation modes enables both biologically faithful and scalable robotic behaviors.

• Fluidic actuation: Pressurized pneumatic or hydraulic networks, including fiber-reinforced bellows and distributed antagonistic chambers, deliver large deflections and force outputs. Models typically relate pressure inputs to curvature via $κ = M/(EI)$, where $M$ is generated by chamber pressure, $E$ the Young's modulus, and $I$ the sectional inertia (Wang et al., 2021, Zhang et al., 2022).
• Dielectric elastomer actuation: DEAs convert Maxwell stress ($σ_M = ε_0 ε_r E^2$) into bending or extension for high-speed, low-inertia swimming and flapping (Zhang et al., 2023), achieving >1 BL/s in compact manta-inspired robots.
• Distributed artificial muscles: Axial placement and independent control of HASELs in fish bodies enables traveling wave formation, higher thrust at elevated resonant modes, and embedded kinematic feedback (Soto et al., 15 Apr 2025).
• Biohybrid and chemical actuation: Autonomous gas generation, combustion-driven pulses, and even integrated living muscle tissue have been harnessed for swimming, gliding, and leaping behaviors (Bonofiglio et al., 2023, He et al., 2023).
• Origami transform mechanisms: Single-DOF, force-amplifying folding architectures enable rapid, robust shape changes for gripping, crawling, or volume modulation (Guo et al., 14 Mar 2025, Singh et al., 13 Nov 2025).
• Adhesion actuators: Positive-pressure, doming strategies with bilayer constructs deliver robust, reversible grip on surfaces in both air and under water, supporting amphibious climbing robots (Tang et al., 2018).

4. Sensing, Control, and Intelligent Behavior

Advanced underwater soft robots integrate proprioceptive and exteroceptive sensing with open-loop, closed-loop, and learning-based control strategies.

• Embedded strain and curvature sensing: Liquid-metal, carbon-nanotube, or foil strain sensors are laminated or molded into soft spines, enabling real-time midline reconstruction and feedback for gait optimization (Soto et al., 15 Apr 2025, Wang et al., 18 Jan 2026).
• Hydrodynamic and tactile perception: Lateral-line analogs, fiber Bragg gratings, and soft tactile arrays monitor flow, contact, or object properties, facilitating embodied interaction and adaptive manipulation (Li et al., 16 Aug 2025).
• Closed-loop model and model-free control: Nonlinear model predictive control (NMPC) addresses wave-induced disturbances on soft manipulators, yielding up to 84% error reduction in set-point/trajectory tracking tasks (Walker et al., 2024). Bang-bang logic and fluidic valves provide low-power, electronics-free feedback in gliders (Bonofiglio et al., 2023).
• Embodied intelligence: Distributed actuation and compliance enable robots to passively negotiate obstacles, exploit environmental forces, and simplify control requirements—seen in flagellum-driven redundancy, passively morphing appendages, and quick-return mechanisms (Mathew et al., 25 Mar 2025, Zhang et al., 2024).
• Learning and optimization: Reinforcement learning and multi-criteria decision frameworks (G1-VIKOR) are used to automate design-choice mapping from biological behaviors to robotic hardware (Chen et al., 16 Apr 2025).

5. Locomotion Modes, Manipulation, and Functional Capabilities

Soft robots achieve a wide range of underwater behaviors, spanning propulsion, adhesion, object grasping, and environment-adaptive transformations.

Locomotion Modes:

Bio-inspiration	Robotic Implementation	Performance Metrics
Fish (BCF/MPF)	Modular elastomer fish; DEA/HASEL fins; phasic drives	0.25–0.81 BL/s (Zhang et al., 2023, Wang et al., 18 Jan 2026)
Jellyfish	Bell-actuated pulsed swimmers	0.8 mm/s (IPMC), up to 300 m depths (Wang et al., 18 Jan 2026)
Cephalopod	Jetting actuators: elastic mantles, combustion	Up to 1100 N peak thrust; 850 BL/s² accel (He et al., 2023)
Snake/Eel	Soft continuum robot snakes, McKibben or SMA actuation	Smooth, undulatory motion; modeling only (Godage, 2019)
Flagella	Dodecahedral flagella drones (ZodiAq, BactoBot)	0.04–0.06 m/s; omnidirectional, safe interaction (Chowdhury et al., 25 Sep 2025, Mathew et al., 25 Mar 2025)
Amphibious Climbing	Spiral-channel adhesion, doming actuators	≥200 g load, 1.6 BL/min climbing speed (Tang et al., 2018)

Manipulation and Grasping:

• Anthropomorphic multi-chamber water-hydraulic hands achieve forehand/backhand grasps under high pressure (up to 10 N five-finger force) with rapid response and compliance (Wang et al., 2021).
• Origami-based single-DOF fish-mouth grippers employ geometric closure for compliant, delicately adjustable grasping and scooping, with high success rates even on soft marine targets (Guo et al., 14 Mar 2025).
• Distributed soft tentacles and suction pads conform to substrates for both manipulation and reversible adhesion (Tang et al., 2018, Li et al., 16 Aug 2025).

Multi-environment and Adaptive Morphing:

• PuffyBot and related projects achieve seamless switching between terrestrial crawling, aquatic swimming, hovering, and submergence/resurfacing via shape-morphing scissor mechanisms and volume modulation (Singh et al., 13 Nov 2025).

6. Environmental Integration, Applications, and Challenges

Bio-inspired soft robots are uniquely adapted for underwater environments due to their compliance, minimal noise generation, and low harm to biological systems.

• Applications: Inspection and maintenance (ship hulls, aquaculture nets), ecological surveys, window cleaning, payload delivery, sampling fragile biota (corals, jellyfish), distributed environmental monitoring via low-cost swarms (Li et al., 16 Aug 2025, Singh et al., 13 Nov 2025).
• Eco-friendliness: Fluidic, dielectric and passive actuation regimes reduce acoustic footprints and physical disturbance, enabling interaction with sensitive fauna and habitats (Wang et al., 18 Jan 2026).
• Endurance and autonomy: Modular gliders with purely mechanical fluidic circuits demonstrate >150 m travel per CO₂ charge at 28 mW/m; soft glider and swimmer designs support distributed, disposable units for broad environmental coverage (Bonofiglio et al., 2023, Chowdhury et al., 25 Sep 2025).
• Key limitations: Energy density (battery and actuation), durability in deep sea (elastomer fatigue, aging), communications (acoustic and magnetic constraints), and integration of robust, multi-modal sensing. Autonomous navigation and manipulation in cluttered, high-disturbance environments remain open challenges (Wang et al., 18 Jan 2026, Li et al., 16 Aug 2025, Singh et al., 13 Nov 2025).

7. Prospects and Design Guidelines

Emerging directions encompass:

• Hybridization of multiple locomotion and manipulation principles (BCF-MPF, undulation+jetting, tactile adhesion).
• Development of advanced, tunable, and self-healing material systems (dynamic crosslinkers, composite hydrogels).
• Freeform reversible embedding (FRE) and multi-material 3D printing for integrated actuator–sensor–structure modules.
• Adoption of “biouniversal” paradigms that abstract convergent solutions across taxa (log-spiral appendages, modular skeletons) for greater adaptability (Li et al., 16 Aug 2025).
• Fusion of closed-loop model-based and data-driven control, onboard sensory intelligence, and decentralized swarm coordination.

Underwater bio-inspired soft robots thus represent a rigorously grounded, rapidly expanding research frontier. By leveraging nature's morphological, material, and behavioral paradigms, and by integrating advances in soft actuation, sensing, and computation, these systems are well positioned to enable transformative advances in marine science, technological innovation, and the direct testing of biological hypotheses (Wang et al., 18 Jan 2026, Li et al., 16 Aug 2025, Tang et al., 2018).