Soft robotics has steadily evolved beyond pneumatic and hydraulic actuation, incorporating new control schemes and multifunctional materials to achieve movements and tasks inaccessible to rigid systems. Among these innovations, magnetism has emerged as a versatile tool, enabling tetherless actuation and programmable deformation through the integration of magnetic particles into elastomers. This approach leverages the inherent flexibility of polymers while adding remote control capabilities via magnetic fields.
Magnetic elastomers (MEs) are elastic composites embedded with either hard magnetic particles (hMEs) or soft magnetic particles (sMEs). Hard magnets, such as NdFeB, retain magnetization without an external field and require significant coercivity to demagnetize. Soft magnets, like Fe₃O₄, are easily magnetized and demagnetized, offering low remanence. The microstructure of MEs can be isotropic, with randomly oriented particles, or anisotropic, where particles are aligned during curing under a magnetic field. Anisotropic alignment enhances responsiveness, yielding higher shear stresses and stronger magnetic attraction forces.
Manufacturing methods range from traditional casting in silicone or polyurethane matrices to ultraviolet curing and masked stereolithography (MSLA) 3D printing. Curing under a magnetic field produces particle chains aligned along the field direction, while curing without a field results in random orientation. Scanning electron microscopy, energy-dispersive X-ray spectroscopy, and nano-computed tomography have validated these internal structures.
Magnetorheological elastomers (MREs), a subset of MEs, can change both shape and mechanical properties under a magnetic field. They have been extensively studied for vibration damping and control, with applications including prosthetic springs, tunable isolators, and variable stiffness mounts. Additives such as carbon nanotubes or carbon black can enhance damping performance, while particle volume fraction and anisotropy influence storage modulus and shear behavior.
Hard magnetic elastomers have demonstrated particular promise in soft robotics. By embedding NdFeB particles into elastomers and programming magnetic domains during fabrication, researchers have achieved complex, repeatable deformations—rolling, twisting, folding—under controlled magnetic fields. Millimeter-scale hME-based robots have been shown to swim, rotate, and navigate confined spaces, including vascular networks, opening pathways for minimally invasive medical tools. Wearable hME skins have enabled gesture control and eye-tracking, while origami-inspired designs have incorporated permanent magnets for foldable, remotely actuated structures.
Soft magnetic elastomers offer different advantages, particularly in tunable stiffness and magnetostrictive actuation. Fe₃O₄-based sMEs have been used in laparoscopic manipulators with electromagnet-controlled joints, micro-actuators with reversible bending, and swimming microrobots driven by helical flagella. sME-based grippers have combined magnetic positioning with thermal actuation, while inchworm-inspired robots have employed sME segments for gait control. Valving systems and peristaltic pumps using sMEs demonstrate their potential in fluid handling within soft robotic architectures.
Sensing applications have also emerged. sME skins paired with Hall effect sensor arrays can detect touch and deformation, with neural networks interpreting contact states. Cilia-inspired composites using magnetized nanowires have been explored for tactile sensing. Hybrid sMEs combining magnetic particles with liquid metal have shown resistivity changes under mechanical strain and magnetic fields, enabling selective Joule heating.
Future directions point toward integrating both hard and soft magnetic particles within a single elastomer to create deformable electromagnets or electropermanent magnets. Such designs could generate magnetic fields internally, reducing reliance on external sources and enabling local control for self-sensing, magnetic coupling, and targeted deformation. Flexible solenoids embedded in elastomers could measure changes in remanent magnetism during motion, feeding data into machine learning systems for state estimation.
Another promising avenue is exploiting the interplay between mechanical deformation and magnetic properties. Adjusting particle orientation through strain could modulate magnetic forces, enabling novel actuation modes. Characterizing these dynamic magnetic responses remains an open challenge, with most current studies focusing on static field effects. Comprehensive comparisons of magnetic properties across different ME compositions would inform more deliberate material selection for specific soft robotic functions.
Magnetic elastomers are already enabling remote actuation, damping, deformation, and sensing in soft robotics. Advances in fabrication, domain programming, and hybrid material design continue to expand their capabilities, bringing the vision of fully soft, untethered, and multifunctional robots closer to reality.