Soft Robotic Gripper Mimics Chameleon’s Grip, Boosts Adaptability in Collaborative Robots
In a breakthrough development that could reshape the future of robotic manipulation, a team of engineers from Yantai University and the Chinese Academy of Sciences has unveiled a novel soft gripper inspired by the grasping mechanics of a chameleon’s foot. The newly designed flexible gripper, developed by Hou Zhigang, Zhang Yunhai, Hu Bin, Fang Baosheng, Xu Congwang, and Cui Huafei, leverages the natural elasticity of thermoplastic polyurethane (TPU) to achieve high adaptability, enabling collaborative robots to grasp a wide range of objects—delicate or irregularly shaped—without causing damage.
Published in the International Journal of Advanced Manufacturing Technology, the research introduces a gripper that combines biomimicry, advanced materials science, and computational design to overcome longstanding challenges in robotic end-effector technology. Unlike traditional rigid grippers that rely on high-pressure contact and risk damaging fragile items, this new design adapts its shape to the object it grasps, forming a gentle, conforming envelope around it—much like a chameleon’s toes curl around a twig.
The inspiration came from observing how chameleons use their opposable, soft-tipped digits to grip branches of varying diameters and textures. “We noticed that nature often solves complex engineering problems with elegant simplicity,” said Hou Zhigang, lead author and associate professor at Yantai University’s School of Electromechanical and Automotive Engineering. “The chameleon doesn’t need sensors or feedback loops to adjust its grip—it just works. We wanted to replicate that instinctive adaptability in a robotic system.”
The result is a two-fingered, opposable gripper fabricated entirely from TPU using 3D printing technology. TPU was selected for its unique combination of elasticity, durability, and mechanical resilience. Unlike rubber, which can degrade under repeated stress, or PVC, which lacks sufficient tensile strength, TPU offers a balanced profile of high elasticity and abrasion resistance. It can be stretched, bent, and compressed repeatedly without permanent deformation, making it ideal for repeated gripping cycles in industrial and service robotics.
What sets this design apart is not just the material but the internal architecture. The researchers employed topology optimization—a computational method that redistributes material within a given design space to maximize performance under specific constraints. Using SOLIDWORKS Simulation, they began with a solid block of virtual material and applied algorithmic stress simulations to determine where structural support was necessary and where mass could be removed without compromising strength.
The optimization process retained the outer contour of the gripper while carving out an intricate internal lattice of thin, flexible ribs. These ribs act as passive springs, allowing the fingers to deform under load while maintaining structural integrity. The final design measures 100 mm in length, 50 mm in width, and 50 mm in thickness, with a central mounting surface that interfaces directly with a standard electric linear actuator.
“The beauty of topology optimization is that it lets the physics guide the design,” explained Zhang Yunhai, a graduate researcher involved in the project. “Instead of relying on intuition or legacy designs, we let the software tell us where material is needed. This not only reduced weight but also improved the gripper’s responsiveness and energy efficiency.”
Once the optimized model was finalized, the team conducted finite element analysis (FEA) using WELSIM, a specialized software for structural simulation. The analysis focused on stress distribution and displacement under a 10-newton actuation force—typical of collaborative robot applications where safety and precision are paramount.
The simulations revealed a maximum stress of just 0.373 megapascals, well below TPU’s tensile strength of 55 MPa, indicating a large safety margin. More importantly, the fingers exhibited a maximum displacement of 12.5 millimeters, demonstrating significant deformation capability. This large deflection allows the gripper to wrap around objects of varying geometries, from cylindrical cans to irregularly shaped tools, ensuring a secure and damage-free hold.
To explore the limits of flexibility, the researchers also tested a modified version of the gripper with no internal support structure—essentially a hollow, flexible shell. In this configuration, the displacement increased to 13 millimeters, and the stress distribution remained within safe limits. While less robust than the optimized version, this “unsupported” design proved capable of handling extremely delicate items such as empty aluminum cans, paper boxes, and plastic cups—objects that would typically collapse under the grip of conventional robotic hands.
“This shows that we can tune the gripper’s compliance simply by adjusting the internal structure,” noted Hu Bin, an assistant researcher at the Institute of Automation, Chinese Academy of Sciences. “For heavy-duty tasks, we can add more support. For fragile objects, we can reduce it. It’s a scalable, modular approach to soft robotics.”
To validate the design in real-world conditions, the team mounted the gripper on a Baxter collaborative robot—a dual-arm platform developed by Rethink Robotics known for its safety, ease of programming, and human-robot interaction capabilities. The experiments involved picking up and placing eight diverse objects: a computer mouse, a glasses case, a 3D-printed toy, a metal can, a pair of scissors, a small garden trowel, a pair of pliers, and a fresh apple.
Each object varied in weight (from 50 to 350 grams), material composition (plastic, metal, wood, fabric, fruit), and geometry. Despite these differences, the gripper successfully picked up every item on the first attempt, with no slippage or damage. The soft TPU surface conformed naturally to curved, angular, and uneven surfaces, distributing contact pressure evenly and eliminating stress concentrations.
“The real test was the apple,” said Xu Congwang. “It’s soft, irregular, and easily bruised. But the gripper handled it like a human hand—gently, securely, and without leaving a mark. That’s the kind of performance we’re aiming for in food handling, medical devices, and consumer electronics assembly.”
In a second round of tests, the unsupported version of the gripper was evaluated against even more fragile items. It successfully grasped a crumpled cigarette box, a thin-walled plastic cup filled with water, and an empty soda can—all without crushing or puncturing them. This level of sensitivity opens up applications in recycling facilities, where robots must sort lightweight, deformable packaging, or in elderly care settings, where robots assist with daily tasks involving delicate household items.
The team also conducted load tests to quantify the gripper’s lifting capacity under different actuation forces. Results showed a nonlinear relationship between input force and maximum payload. At low forces (10–20 N), the unsupported gripper outperformed its supported counterpart, thanks to its superior flexibility and surface conformity. However, as the actuation force increased beyond 30 N, the supported version demonstrated higher load capacity, capable of lifting over 1 kilogram—nearly three times the weight of the heaviest test object.
“This suggests a trade-off between gentleness and strength,” observed Cui Huafei. “For light, fragile objects, the unsupported design is ideal. For heavier items, especially those with smooth surfaces that require more friction, the structured version is better. The key is matching the gripper design to the task.”
One of the most significant advantages of this design is its simplicity. Unlike many soft robotic grippers that rely on pneumatic or hydraulic actuation—requiring compressors, valves, and tubing—this gripper is driven by a standard electric linear actuator. This eliminates the need for complex auxiliary systems, reduces maintenance, and improves energy efficiency. It also makes integration with existing robotic platforms straightforward.
“Pneumatic soft grippers are powerful, but they’re bulky and noisy,” said Hou Zhigang. “By using electric actuation and passive compliance, we’ve created a gripper that’s quiet, compact, and easy to deploy. It’s a practical solution for real-world automation.”
The implications for industry are substantial. In manufacturing, where robots increasingly work alongside humans, safety and adaptability are critical. Rigid grippers can injure workers or damage products, especially in dynamic environments where object positions vary. Soft grippers mitigate these risks, enabling safer human-robot collaboration.
In logistics and warehousing, where robots handle everything from cardboard boxes to glass bottles, the ability to grasp diverse objects without retooling saves time and increases throughput. Amazon and other e-commerce giants have already invested heavily in robotic picking systems, but current solutions struggle with soft or irregular items. A gripper like this could significantly improve pick rates and reduce product damage.
In agriculture, where harvesting robots must handle fruits and vegetables without bruising them, soft grippers are essential. Previous designs, such as underactuated mechanisms or suction-based systems, often fail on wet or uneven surfaces. The chameleon-inspired gripper, with its conforming surface and high friction, offers a more reliable alternative.
Even in healthcare, where robots assist in surgery or patient care, gentle manipulation is crucial. Instruments must be handled precisely, and patients must be treated with care. A soft gripper that mimics biological motion could enhance the safety and dexterity of medical robots.
The research also highlights the growing role of biomimicry in robotics. Nature has spent millions of years refining solutions to mechanical challenges, from locomotion to manipulation. By studying biological systems, engineers can discover principles that are difficult to derive through conventional design.
“The chameleon’s foot is a masterpiece of evolutionary engineering,” said Zhang Yunhai. “It’s lightweight, energy-efficient, and self-adapting. We’re not copying it exactly—we’re abstracting its principles and applying them to a new context. That’s the essence of biomimetic design.”
The team emphasized that their gripper is not a one-size-fits-all solution but a platform for further innovation. By adjusting parameters such as wall thickness, internal rib density, and overall dimensions, the same basic design can be customized for different applications. Future work may explore multi-material printing to create gradients in stiffness, or integrate embedded sensors for real-time feedback.
Another promising direction is scalability. The current design is optimized for small to medium objects, but the same principles could be applied to larger grippers for industrial handling or micro-scale versions for lab automation.
“The beauty of this approach is its versatility,” said Hu Bin. “We’re not just building a better gripper—we’re building a design methodology that can be adapted across industries and scales.”
The paper has been well received by the robotics community, particularly for its practical focus and rigorous validation. By combining simulation, material science, and real-world testing, the authors have demonstrated a clear path from concept to application.
As collaborative robots become more prevalent in workplaces, homes, and hospitals, the demand for intelligent, safe, and adaptable end-effectors will only grow. This chameleon-inspired gripper represents a significant step forward—proving that sometimes, the best engineering solutions come not from complex algorithms or expensive hardware, but from observing the natural world.
Soft Robotic Gripper Mimics Chameleon’s Grip, Boosts Adaptability in Collaborative Robots
Hou Zhigang, Zhang Yunhai, Hu Bin, Fang Baosheng, Xu Congwang, Cui Huafei, Yantai University and Chinese Academy of Sciences, International Journal of Advanced Manufacturing Technology, DOI: 10.16731/j.cnki.1671-3133.2021.05.008