1 Fundamentals of Soft Robotics
1.1 Definition and Scope
Soft robotics is a subfield of mechanical engineering and robotics that emphasizes the use of highly compliant, deformable materials—such as elastomers, gels, and textiles—to create robotic systems capable of safe, adaptive, and biologically inspired movement. Unlike conventional rigid robots, soft robots leverage material compliance to interact gently with humans, navigate unstructured environments, and perform tasks like gripping delicate objects or crawling through confined spaces. The scope of soft robotics encompasses material science, novel actuation methods, continuum mechanics modeling, sensing, and control, with applications spanning medicine, search-and-rescue, and wearable assistive devices.
1.2 Historical Development
The roots of soft robotics trace back to early pneumatic artificial muscles developed in the 1950s, such as the McKibben muscle. However, the field gained formal recognition in the early 2000s, driven by advances in elastomeric fabrication and the need for safer human-robot interaction. Landmark achievements include the development of the first entirely soft autonomous robot (the "Octobot") in 2016 and the emergence of fully soft grippers and walkers. The field continues to evolve rapidly, leveraging insights from biology and materials science.
2 Materials for Soft Robotics
2.1 Elastomers and Polymers
Elastomers and polymers are the most common base materials for soft robots due to their high stretchability and low stiffness. They can be molded, cast, or 3D-printed into complex shapes, often incorporating internal channels or chambers for actuation.
2.1.1 Silicones
Silicones, such as polydimethylsiloxane (PDMS) and Ecoflex, are widely used for their biocompatibility, chemical stability, and ability to undergo large deformations without permanent damage. They are the material of choice for many fluidically actuated soft robots, including grippers and biomedical devices.
2.1.2 Hydrogels
Hydrogels are water-swollen polymer networks that mimic biological tissues. They can change volume in response to stimuli (e.g., pH, temperature, electric fields), making them suitable for soft actuators and sensors. However, their mechanical strength and durability are generally lower than those of silicones.
2.2 Smart Materials
Smart materials respond to external stimuli—such as heat, electric fields, or magnetic fields—with a change in shape or stiffness, enabling direct actuation or variable compliance.
2.2.1 Shape Memory Alloys
Shape memory alloys (SMAs), like Nitinol, can be pre-deformed and then return to a remembered shape when heated. In soft robotics, thin SMA wires or springs embedded in elastomers serve as tendon-like actuators, capable of generating large forces with simple electrical control.
2.2.2 Dielectric Elastomers
Dielectric elastomers (DEs) are soft, insulating polymers that expand in area when subjected to a high electric field. They can be used as artificial muscles, offering high strains and fast response times, though they require high voltages and careful insulation.
2.3 Textiles and Fabrics
Textiles and woven fabrics provide a lightweight, flexible substrate for soft robotics. They can be coated or impregnated with responsive materials (e.g., shape memory polymers, conductive threads) to create wearable exosuits, soft sensors, or inflatable structures. Their breathability and comfort make them ideal for human-interactive applications.
3 Actuation Principles
3.1 Fluidic Actuation
Fluidic actuation uses compressed air or liquid to pressurize internal cavities in a soft structure, causing the robot to bend, extend, or twist.
3.1.1 Pneumatic Artificial Muscles
Pneumatic artificial muscles (PAMs), such as the McKibben muscle, consist of a flexible tube with a braided shell. When pressurized, they contract, mimicking biological muscle contraction. PAMs offer high power-to-weight ratios and intrinsic compliance.
3.1.2 Hydraulic Actuation
Hydraulic actuation uses incompressible fluids (e.g., water, oil) to achieve higher forces and more precise control than pneumatics. Soft hydraulic actuators are often used in underwater robots and medical devices where fluid leakage is tolerable.
3.2 Tendon-Driven Actuation
Tendon-driven systems use cables or threads routed through a soft body. When tension is applied (by a motor or actuator), the robot bends or contracts. This approach offers simplicity and precise force transmission, but requires a rigid or semi-rigid anchor point for the motor.
3.3 Thermal and Electrical Actuation
Thermal and electrical methods alter the material properties of soft components to produce motion.
3.3.1 Shape Memory Polymers
Shape memory polymers (SMPs) can be programmed to recover a pre-defined shape when heated beyond a transition temperature. They are lightweight and can be combined with embedded heaters for local actuation, though response times are typically slower than for SMAs.
3.3.2 Electroactive Polymers
Electroactive polymers (EAPs) deform under an electric field. They include dielectric elastomers (described earlier) and ionic polymer-metal composites (IPMCs), which bend when an electric field drives ion movement. EAPs enable low-voltage actuation but often suffer from low force output or complex fabrication.
4 Modeling and Control
4.1 Kinematic and Dynamic Models
Modeling soft robots is challenging due to their infinite degrees of freedom and nonlinear material behavior. Two main approaches are used.
4.1.1 Continuum Mechanics Approaches
Continuum models treat the robot as a deformable body described by partial differential equations (e.g., Cosserat rod theory, finite element methods). These models capture detailed strain distributions but are computationally intensive, limiting their use for real-time control.
4.1.2 Lumped Parameter Models
Lumped parameter models approximate the robot as a series of discrete elements (e.g., masses, springs, dampers) or as a backbone with piecewise constant curvature. These simplified models trade accuracy for computational speed, making them suitable for real-time control and simulation.
4.2 Sensing and Feedback
4.2.1 Soft Sensors
Soft sensors measure strain, pressure, or curvature using materials that change electrical resistance, capacitance, or light transmission. Common examples include conductive silicone (e.g., carbon-black-filled elastomers), liquid-metal-filled microchannels, and fiber optic sensors. They must be stretchable, durable, and seamlessly integrated with the robot body.
4.2.2 Control Strategies
Control of soft robots is complicated by hysteresis, nonlinearity, and slow dynamics. Two broad strategies are employed.
4.2.2.1 Open-Loop Control
Open-loop control relies on precomputed actuation sequences, often derived from empirical calibration or simulation. It is simple and works well for repetitive tasks in known environments, but lacks adaptability to disturbances or unknown objects.
4.2.2.2 Closed-Loop Control
Closed-loop control uses sensor feedback to adjust actuation in real time. Common techniques include PID control, model predictive control, and learning-based methods (e.g., reinforcement learning). Feedback enables robust performance despite material uncertainties, but requires reliable, low-latency soft sensors.
5 Applications
5.1 Biomedical Devices
The inherent compliance of soft robots makes them ideal for safe interaction with delicate biological tissues.
5.1.1 Surgical Tools
Soft robotic tools, such as flexible endoscopes and steerable needles, can navigate through tight, tortuous anatomical pathways without causing trauma. Pneumatically actuated soft grippers are used for atraumatic tissue manipulation in laparoscopic surgery.
5.1.2 Prosthetics and Orthotics
Soft prosthetic hands and wrists offer lightweight, compliant grasping that adapts to object shape. Soft orthotic exosuits (see Section 5.3) can assist with rehabilitation or compensate for muscle weakness.
5.2 Locomotion and Manipulation
5.2.1 Soft Grippers
Soft grippers are among the most commercially successful soft robotic products. They consist of inflatable fingers that can conform to a wide variety of objects—from fragile fruit to irregularly shaped industrial parts—without sophisticated sensing or control.
5.2.2 Walking and Crawling Robots
Soft robots that walk, crawl, or swim use actuation patterns to generate undulatory or peristaltic motion. Examples include the soft quadruped "Mochibot" and the soft crawler "Meshworm". These robots excel in rugged, sandy, or fluid environments where rigid robots would get stuck.
5.3 Wearable Robotics
5.3.1 Exosuits
Soft exosuits are textile-based wearable devices that apply forces to the body to augment human movement. They are typically actuated by cables or pneumatics and are lighter and more comfortable than rigid exoskeletons. Applications include reducing metabolic cost during walking and assisting with lifting tasks.
5.3.2 Assistive Devices
Soft assistive devices help individuals with motor impairments perform activities of daily living. Examples include soft gloves that aid hand opening/closing and ankle-foot orthoses that prevent foot drop. Their softness ensures user comfort and safety during prolonged use.
6 Challenges and Future Directions
6.1 Fabrication and Scalability
Current fabrication methods, such as silicone casting and layer-by-layer assembly, are labor-intensive and difficult to scale. Future production may rely on automated 3D printing, molding with embedded channels, and textile manufacturing techniques to enable cost-effective mass production.
6.2 Durability and Reliability
Soft materials are prone to tearing, fatigue, and degradation under repeated loading or harsh conditions. Improving material toughness, self-healing capabilities, and protective coatings is critical for long-term real-world use.
6.3 Integration with Rigid Components
Many soft robots still require rigid components—such as pumps, valves, and batteries—for power and control. Fully soft systems demand advances in flexible electronics, soft batteries, and embedded fluidic logic to eliminate rigid interfaces.
6.4 Autonomy and Power Sources
Soft robots often rely on tethered power or bulky external pumps, limiting their autonomy. Developing integrated soft energy sources (e.g., chemical reactions, compressed gas stored in soft bladders) and on-board soft controllers (e.g., pneumatic logic circuits) are active research directions for untethered operation.