Soft robotics represents a paradigm shift in machine design, moving away from rigid joints and electric motors toward structures made of compliant materials—elastomers, gels, and textiles—that can bend, stretch, and conform to their environment. These robots excel in tasks where safety, adaptability, and delicate interaction are critical, such as minimally invasive surgery, assistive wearables, and handling fragile objects in manufacturing or agriculture. However, the complexity of soft robotic systems—which often require integrated actuators, sensors, and fluidic channels—has historically outpaced available fabrication techniques. Recent breakthroughs in additive manufacturing, microfabrication, and bio-inspired processes are now closing that gap, enabling engineers to produce intricate, multifunctional soft structures with unprecedented precision and reproducibility. This article surveys the most promising emerging fabrication techniques, their underlying principles, current limitations, and the transformative potential they hold for the field.

Advanced Additive Manufacturing for Soft Robotics

Three‑dimensional printing has become a cornerstone of soft robotic fabrication because it allows direct, layer‑by‑layer construction of complex geometries without the need for molds or tooling. Unlike conventional 3D printing with rigid thermoplastics, soft robotics demands flexible, elastomeric materials that can undergo large deformations while maintaining structural integrity. Several specialized additive manufacturing techniques have emerged to meet these requirements.

Direct Ink Writing of Elastomers

Direct ink writing (DIW) is an extrusion‑based 3D printing process that deposits a viscoelastic ink through a nozzle onto a build platform. For soft robotics, inks are typically composed of silicone elastomers, polyurethanes, or hydrogels with carefully tuned rheology to ensure continuous flow and immediate shape retention. Researchers have used DIW to fabricate pneumatic actuators with built‑in channels, soft grippers with variable stiffness, and even complete soft robots in a single print job. A key advantage of DIW is the ability to embed sacrificial materials—such as wax or Pluronic F‑127—that can later be removed to create hollow channels, enabling complex fluidic networks that would be impossible to mold. Recent work at Harvard’s Soft Robotics Lab has demonstrated DIW printing of silicone‑based actuators with integrated sensing chambers, achieving both bending and elongation in a single monolithic structure.

Multi‑Material 3D Printing

True soft robotic systems often require regions of different stiffness, color, or conductivity within the same component. Multi‑material 3D printing addresses this by switching between multiple print heads or by using a single nozzle fed from several reservoirs. The Stratasys PolyJet system can deposit droplets of photopolymerizable elastomers with varying durometers, while custom‑built multi‑material DIW printers can co‑print conductive carbon‑filled silicone for stretchable sensors alongside insulating elastomers. This capability enables the fabrication of soft robots with embedded strain gauges, capacitive touch sensors, or even dielectric elastomer actuators—all printed in one pass. The major challenge remains the adhesion between different materials; dissimilar elastomers can delaminate under cyclic loading, though advances in surface treatment and ink chemistry are steadily improving bond strength.

Embedding Functional Components During Print

Beyond printing passive structures, additive manufacturing now allows the direct integration of active components—microfluidic valves, shape‑memory alloy wires, or even miniature battery cells—as the robot is built. In a technique known as embedded 3D printing, a nozzle deposits elastomer around a pre‑placed functional element, locking it in place. For example, researchers at the EPFL Soft Transducers Laboratory have printed soft pneumatic actuators with embedded optical fibers that simultaneously actuate and sense deformation, eliminating the need for post‑assembly. This integration reduces part count, increases reliability, and shortens development cycles.

Soft Lithography and Microfabrication

When sub‑millimeter precision is required—for instance, in micro‑grippers or endovascular robots—soft lithography remains the method of choice. Originally developed for microelectronics and microfluidics, soft lithography has been adapted to produce soft robotic components with feature sizes down to a few micrometers.

Replica Molding and Stamping

The classical soft lithography process begins with a master mold created using photolithography or micromachining. A liquid elastomer (typically polydimethylsiloxane, PDMS) is poured over the mold, cured, and peeled off to yield a flexible replica. This technique is ideal for fabricating thin‑film actuators, microfluidic valves, and pneumatic networks with intricate channel geometries. Recent improvements in two‑photon polymerization allow the creation of masters with truly three‑dimensional microfeatures, enabling molds for soft robotic structures that incorporate overhangs, undercuts, and vertical channels—features impossible with traditional SU‑8 photoresist. The resulting parts can achieve aspect ratios exceeding 20:1, which is critical for high‑displacement actuators.

Microfluidic Soft Actuators

Soft lithography excels at producing microfluidic channels that act as artificial muscles when pressurized. The PneuNet (pneumatic network) actuator, a staple of soft robotics, is typically fabricated by replica molding a series of expanding chambers connected by a central channel. Advances in soft lithography have led to finer chamber geometries, higher density of actuation elements, and the integration of embedded check valves that enable sequential inflation without external controllers. Researchers have also combined soft lithography with 3D‑printed molds to produce hybrid actuators that combine macroscopic bending with microscopic surface textures for improved grip.

Scalability and Throughput

One persistent criticism of soft lithography is its limited scalability—each mold produces only one design, and manual alignment of layers can be labor‑intensive. However, roll‑to‑roll processing and automated alignment systems are emerging. For example, the use of UV‑curable elastomers in a reel‑to‑reel format allows continuous fabrication of soft robotic films with embedded microchannels, dramatically increasing production speed. This technique is particularly promising for wearable soft robots, where large sheets of actuating material must be manufactured at low cost.

Bio‑Inspired and Self‑Assembling Fabrication

Nature builds soft structures not by assembly lines but through self‑organization, growth, and adaptation. These principles are inspiring a new class of fabrication methods that reduce external intervention and enable structures with built‑in homeostasis.

Self‑Assembly of Soft Matter

Self‑assembly leverages the inherent tendency of certain materials to organize into ordered structures under specific conditions. In soft robotics, block copolymers, liquid crystal elastomers, and DNA‑based hydrogels can be programmed to form micro‑actuators, porous scaffolds, or even networks of channels that mimic vascular systems. For instance, liquid crystal elastomers (LCEs) can be synthesized to self‑align into layers that undergo large, reversible shape changes when heated or exposed to light. By patterning the alignment director via surface treatment or magnetic fields, researchers can create soft actuators that bend, twist, or undulate without any mechanical input after the initial synthesis. A recent demonstration from the UC Santa Barbara Materials Research Lab showed a self‑assembled LCE ribbon that walks autonomously when placed on a hot surface—a soft robot grown, not assembled.

Growth and Biofabrication

Inspired by biological development, scientists are exploring ways to grow soft robots by depositing material in a controlled manner, similar to how a plant grows new tissue. One approach uses additive manufacturing of living cells (bio‑printing) to create hydrogel‑based robots that can be colonized by muscle cells, resulting in biohybrid actuators that contract in response to electrical or chemical stimuli. Another grows structures from a nutrient bath: a robot arm that extends by polymerizing new material at its tip, akin to the growth of a fungal hypha. These methods are still at the proof‑of‑concept stage but promise robots that can self‑repair, adapt their shape, and even degrade after use.

Hybrid Bio‑Inspired Techniques

Combining self‑assembly with traditional fabrication yields hybrid processes that capture the best of both worlds. For example, a 3D‑printed scaffold can be infused with a self‑assembling gel that creates internal porosity, producing a lightweight yet robust structure with embedded channels for fluidic actuation. Another hybrid approach uses magnetic fields to align ferromagnetic particles within a curing elastomer, creating a soft material with programmable anisotropy. This technique, sometimes called magnetolithography, enables the fabrication of soft robots that respond to magnetic fields with complex motions—without any rigid components.

Multimaterial and Gradient Manufacturing

Soft robotic systems rarely consist of a single homogeneous material. To achieve natural motion, a robot’s body often needs a smooth transition from very compliant to stiffer regions, mimicking the way biological muscle attaches to tendon and bone. Gradient manufacturing methods address this by varying material composition continuously or in discrete steps.

Functionally Graded Elastomers

By controlling the ratio of base elastomer to crosslinker or filler during printing, it is possible to create a part with a continuous stiffness gradient. For example, a soft robotic finger can be printed with a flexible tip that gradually becomes stiffer toward the base, eliminating stress concentration at joints and improving durability. Multi‑nozzle DIW and voxel‑based digital light processing (DLP) are the primary tools for achieving such gradients. Voxel‑based DLP, in particular, allows each volumetric pixel to have its own cure depth and material composition, offering the highest degree of control currently available. A 2024 study published in Science Robotics demonstrated a fully printed soft hand with graded stiffness that could grasp objects ranging from an egg to a weightlifting dumbbell without any control software—purely through passive compliance.

Embedding Rigid Inserts

In some applications, a purely soft structure is insufficient—for instance, when a robot must exert significant force or carry a payload. Emerging fabrication techniques can co‑print soft elastomers with rigid thermoplastics or metal inserts. A seam‑free transition is achieved using a tie‑layer adhesive that gradients from soft to hard. This approach has been used to fabricate soft grippers with rigid mounting flanges, allowing them to be attached to industrial robot arms without post‑production fasteners.

Application Spotlight: Medical and Wearable Soft Robots

The fabrication methods described above are enabling a new generation of soft robots for healthcare and human augmentation. For instance, 4D printing—where a 3D‑printed structure changes shape over time in response to a stimulus—is being used to create stents that open in the body at body temperature. Multimaterial DIW allows the printing of soft orthotic insoles that sense pressure and adjust stiffness in real time. Soft lithography produces micro‑actuators for endoscopes that can navigate the colon without damaging tissue. In each case, the fabrication technique directly determines the device’s performance and safety, making continued innovation essential.

Challenges and Future Outlook

Despite rapid progress, significant hurdles remain. Most emerging fabrication techniques suffer from relatively low throughput compared to injection molding, limiting their use to prototyping and low‑volume production. Material durability under repeated cyclic loading remains a concern—many soft elastomers fatigue quickly, especially when printed or molded with internal stresses. Additionally, the integration of power sources, control electronics, and logic into fully soft bodies is still an active research area. Future directions include the development of self‑healing materials that can be incorporated into fabrication processes, the use of machine learning to optimize print parameters for complex actuator designs, and the creation of modular fabrication platforms that can switch between additive, lithographic, and self‑assembly methods within a single build. Ethical considerations, particularly around biodegradability and the environmental impact of synthetic elastomers, will also shape the materials chosen in the years ahead.

The emerging fabrication techniques for soft robotic structures—3D printing, soft lithography, bio‑inspired self‑assembly, and gradient manufacturing—are more than incremental improvements; they represent a fundamental expansion of what is possible. As these methods mature, they will unlock soft robots that are not only more complex and capable but also manufactured more efficiently and sustainably. For researchers, engineers, and clinicians, staying abreast of these fabrication advances is the key to turning the promise of soft robotics into real‑world impact.