Soft Grip, Heavy Lift: A New Approach to Robotic Handling

Author: Denis Avetisyan

Researchers have developed a hydraulically-powered soft robotic gripper capable of securely grasping objects weighing up to 20kg, opening new possibilities for automated handling of large payloads.

This soft robotic gripper utilizes a hydraulic system-comprising soft, pneumatically-actuated fingers, integrated pressure sensors, an oil reservoir, and a motor-driven pump-to achieve controlled grasping through fluidic manipulation.

This review details the design methodology, combining mathematical modeling, finite element analysis, and closed-loop control for a robust and adaptable soft gripper system.

While most soft robotic grippers struggle with heavy payloads due to limitations of pneumatic actuation, this paper details a design methodology for a hydraulically-driven soft gripper capable of reliably grasping objects weighing up to 20kg, as presented in ‘Design Methodology of Hydraulically-driven Soft Robotic Gripper for a Large and Heavy Object’. By integrating mathematical modeling, finite element analysis, and experimental validation, we demonstrate a system achieving both high grasping force and adaptable manipulation. This approach enables a significant increase in payload capacity compared to pneumatically-actuated counterparts – but how can these designs be further optimized for increasingly complex and variable grasping scenarios?

Beyond Fixed Constraints: Embracing Adaptability in Gripping Systems

Conventional robotic grippers, often relying on fixed pincers or suction cups, frequently falter when confronted with the variability of real-world objects. These designs excel with standardized, consistently shaped items in controlled environments, but struggle with produce, irregularly formed components, or anything lacking a consistent, easily grasped surface. This limitation significantly hinders broader automation efforts, as a substantial portion of potential tasks require handling objects that deviate from ideal geometries or possess fragile exteriors. Consequently, industries seeking fully automated solutions often face bottlenecks where human dexterity remains essential, necessitating a shift towards gripping technologies capable of conforming to, rather than forcing objects to fit, predefined parameters. The inability to reliably grasp diverse items represents a key impediment to realizing the full potential of robotic systems across logistics, manufacturing, and even domestic applications.

The limitations of current robotic grasping technology necessitate a shift towards solutions capable of handling the inherent variability of real-world objects and environments. Traditional grippers, often designed for specific, predictable tasks, struggle with items exhibiting irregular geometries, delicate surfaces, or unexpected weight distributions. Consequently, a robust yet adaptable approach to grasping is critical for expanding the scope of robotic automation beyond highly structured settings. Such systems require not only the strength to secure an object, but also the sensitivity to modulate grip force and conform to diverse shapes without causing damage. This adaptability promises to unlock automation potential in fields like agriculture, logistics, and healthcare, where handling a wide range of unpredictable objects is the norm – ultimately bridging the gap between laboratory precision and the messy reality of everyday tasks.

Soft gripper performance, as measured by payload-to-weight ratio, scales inversely with gripper mass, as detailed in Table 4.

Harnessing Compliance: The Principles of Soft Robotic Design

Soft robotic grippers commonly employ flexible materials such as Nitrile Butadiene Rubber (NBR) to achieve enhanced grasping capabilities. Unlike rigid grippers, these materials allow for conformal contact with objects of varying shapes and sizes, maximizing surface area contact and distributing grasping forces. This conformability improves grasp stability, particularly with delicate or irregularly shaped objects, and reduces the risk of damage. The material properties of NBR, including its elasticity and friction coefficient, contribute to a secure hold without requiring precise positioning or high clamping forces. This approach is especially beneficial when handling fragile items or in applications where object geometry is unpredictable.

Kirigami, the art of folding and cutting paper, inspires soft robotic designs by creating structures with patterned cuts that enable controlled deformation along specific axes. When combined with fiber-constrained design, where strategically embedded fibers limit bending to desired planes, these robots achieve predictable and repeatable motions. The fibers act as hinges, guiding the deformation induced by actuators and preventing unwanted buckling or twisting. This approach allows for precise control over the robot’s bending radius and angle, crucial for tasks requiring dexterity and manipulation of objects with varying geometries. The combination yields structures that are both flexible and structurally stable, offering a high degree of control compared to purely flexible systems.

The PCS (Position, Curvature, and Stretch) model is a kinematic tool used to analyze and optimize the behavior of flexible robotic structures. It simplifies complex deformations by representing the structure’s configuration using only three parameters: position of the neutral axis, curvature, and axial stretch. This allows designers to predict the structure’s end-effector pose based on applied actuator inputs and material properties. The model facilitates inverse kinematics calculations, determining the necessary actuator configurations to achieve a desired pose. Furthermore, the PCS framework enables optimization of structural parameters, such as material distribution and geometry, to maximize performance metrics like workspace volume and force transmission. $\Delta L = L_0 \epsilon$ represents axial stretch, where $\Delta L$ is the change in length and $L_0$ is the original length, and ε is the strain.

The soft gripper's maximum payload decreases as bending angle increases, ranging from <span class="katex-eq" data-katex-display="false">40^\circ</span> to <span class="katex-eq" data-katex-display="false">70^\circ</span>. — The soft gripper’s maximum payload decreases as bending angle increases, ranging from $40^\circ$ to $70^\circ$ .

Validating Performance: Modeling and Simulation for Robustness

Finite Element Method (FEM) analysis is integral to the design process, specifically in evaluating the structural integrity of components fabricated from A5052 aluminum. This material was selected for its high strength-to-weight ratio and corrosion resistance; however, its susceptibility to torsional deflection necessitated detailed FEM simulations. These simulations model stress distribution under anticipated loads, allowing engineers to optimize component geometry – including wall thicknesses and ribbing – to minimize twisting. By virtually testing various designs, FEM identifies potential failure points and informs material distribution strategies, ensuring the robotic fingers maintain positional accuracy and withstand operational stresses without deformation or structural compromise. The process includes applying boundary conditions that represent the mounting points and applied forces, followed by iterative refinement of the design based on the simulation results.

Mathematical modeling of the robotic gripper’s mechanics, incorporating payload estimation, is fundamental to determining its lifting capacity and optimizing hydraulic actuation. This process begins with defining the kinematic and dynamic relationships between the gripper’s joints, links, and the manipulated object. Payload estimation, based on anticipated object weights and center of gravity locations, allows calculation of the required torque at each joint. These torque requirements are then translated into necessary hydraulic cylinder forces and volumes, enabling precise control of finger movement. The model also accounts for gravitational forces, inertial forces due to acceleration, and frictional forces within the system. By accurately predicting these forces, the hydraulic actuation system – including pump capacity, valve timing, and cylinder sizing – can be optimized to deliver sufficient force for stable grasping and manipulation while minimizing energy consumption and ensuring responsiveness.

Closed-Loop Control systems are implemented to maintain desired finger angles by continuously monitoring actual position via integrated sensors – typically encoders or potentiometers – and comparing this data to the commanded position from the mathematical model. Any deviation results in a corrective signal being sent to the hydraulic actuators, adjusting their operation to minimize error and achieve the target angle. This feedback loop is crucial for robust performance in dynamic environments, as it compensates for external disturbances, payload variations, and inherent system nonlinearities. The system’s ability to rapidly and accurately adjust to these changing conditions ensures stable grasping and manipulation, even when faced with unpredictable forces or movements. Precise control is achieved through proportional-integral-derivative (PID) controllers, tuned to optimize response time, stability, and accuracy based on the modeled system dynamics.

Finite element method simulations reveal that stress concentrates at the finger cap and hinge points during grasping under pressures of 2.0 MPa, 3.0 MPa, and 4.0 MPa.

Demonstrating Capabilities: Strength and Adaptability in Practice

Recent advancements in soft robotics have yielded a gripper capable of manipulating surprisingly heavy loads. This innovative design successfully grasped and held a 20 kg object, a feat demonstrating considerable strength for a soft-bodied system. The gripper achieves this through a unique combination of flexible materials and a carefully engineered internal structure, allowing it to conform to the object’s shape while maintaining a secure hold. This capability signifies a departure from traditional rigid robotic grippers, opening possibilities for handling delicate or irregularly shaped items without causing damage – and proving that soft robotics can indeed tackle demanding physical tasks.

The newly developed soft robotic gripper distinguishes itself through an impressive strength-to-weight performance. Weighing just 1.13 kg – excluding the necessary actuation systems – the device boasts a payload capacity of 20 kg. This translates to a payload-to-weight ratio of approximately 17.7, a figure significantly exceeding that of many traditional robotic grippers. Such a high ratio isn’t merely a numerical achievement; it allows for faster movements, reduced energy consumption, and the potential for deployment in weight-sensitive applications, like collaborative robots working alongside humans or aerial manipulation tasks.

This soft robotic gripper transcends the limitations of traditional automation through its remarkable adaptability. Unlike rigid grippers designed for specific shapes and sizes, this design successfully handles a diverse array of objects – from delicate produce to awkwardly shaped industrial parts – without requiring recalibration or specialized tooling. This versatility stems from the inherent compliance of the soft materials, allowing the gripper to conform to an object’s geometry and securely grasp it regardless of surface texture or fragility. Consequently, the technology holds significant promise for streamlining processes across numerous sectors, including agriculture, manufacturing, logistics, and even healthcare, where gentle yet reliable handling is paramount. The potential for broad implementation suggests a future where robotic automation is no longer confined to highly structured environments but can seamlessly integrate into complex, real-world scenarios.

This four-finger soft gripper can reliably lift up to 27 kg at a 40-degree wrap angle and 20 kg at 70 degrees when pressurized to 2.0 MPa.

Charting Future Directions: Expanding the Boundaries of Soft Manipulation

The incorporation of electro-adhesion presents a compelling pathway to significantly improve soft robotic gripper performance, particularly when handling surfaces that pose challenges for traditional methods. This technology utilizes electrostatic forces to create a controllable adhesive bond between the gripper and an object, circumventing the limitations of friction or mechanical clamping on smooth, delicate, or irregularly shaped items. By applying a voltage, the gripper can generate an attractive force, allowing it to securely grasp objects without causing damage or slippage – a crucial capability for applications in fields like agriculture, food handling, and advanced manufacturing. Furthermore, electro-adhesion is easily toggled on and off, enabling rapid and precise manipulation, and potentially allowing the gripper to adapt its grip strength based on the object’s weight and fragility, ultimately broadening the scope of tasks achievable through soft robotics.

The recent strides in soft robotic gripper technology are rapidly establishing these systems as pivotal components within next-generation automation landscapes. Unlike traditional rigid robotic arms, soft grippers offer unparalleled adaptability, allowing them to handle a vastly wider range of objects and navigate complex, unstructured environments. This capability is particularly crucial for industries demanding delicate handling – such as food processing, agriculture, and logistics – but extends to sectors like advanced manufacturing and even surgical assistance. By conforming to object shapes and applying controlled force, soft grippers minimize damage and maximize efficiency, promising a future where automation is more versatile, reliable, and integrated into previously inaccessible workflows. Their inherent compliance also enhances safety, allowing for more effective human-robot collaboration and paving the way for truly intelligent and responsive automated systems.

The future of soft robotic manipulation hinges on synergistic advancements in both material science and control algorithms. Current materials, while flexible, often lack the nuanced responsiveness required for truly adaptable grasping; research into novel polymers, composites, and even bio-inspired designs promises to yield surfaces with dynamically tunable friction and stiffness. However, even the most sophisticated materials are limited without corresponding improvements in control; algorithms must move beyond pre-programmed sequences towards real-time feedback and learning capabilities, allowing the gripper to intelligently respond to unexpected variations in object shape, weight, and surface properties. This iterative process – designing materials that enable more complex control, and developing algorithms that exploit new material capabilities – is poised to unlock a new era of robust and versatile robotic manipulation, extending the reach of automation into previously inaccessible domains.

The hydraulic soft gripper utilizes a finger design with an internal reservoir constructed from an NBR circular sheet, a metal ring, and a base, enabling fluid-driven actuation as detailed in the related figure.

The development of this hydraulically-driven soft gripper exemplifies a systemic approach to robotic design. The research meticulously integrates mathematical modeling, finite element analysis, and experimental validation-a holistic strategy vital for achieving robust performance with heavy payloads. This mirrors Robert Tarjan’s observation: “Program structure is more important than program details.” Just as a well-structured program relies on interconnected components, this gripper’s success stems from the seamless interaction between its analytical foundation and physical realization. Ignoring any single aspect-be it the accuracy of the mathematical model or the precision of the fabrication-would compromise the entire system, highlighting the interconnectedness of design elements and the importance of a unified architectural vision.

Where Do We Go From Here?

This work demonstrates a functional, if predictably limited, hydraulic soft gripper. The achievement of a 20kg payload is noteworthy, yet the system’s boundaries-the interfaces between fluid power, compliant structure, and control algorithm-reveal inherent vulnerabilities. Systems break along invisible boundaries-if one cannot see them, pain is coming. Future iterations will undoubtedly focus on increasing payload capacity, but a more fruitful path lies in understanding how this compliance fails. Finite element analysis, while useful, is still a simplification; the real world introduces complexities of material hysteresis and unpredictable deformation that models struggle to capture.

The presented closed-loop control, while functional, remains reactive. True adaptability demands a predictive element-an anticipation of load distribution and grasping instability. This requires not merely sensing deformation, but modeling the object being grasped – its center of mass, fragility, and potential failure modes. Such a system would move beyond simply reacting to slip, and begin to preempt it.

Ultimately, the field must confront the inherent trade-off between compliance and precision. Increasing the number of actuators and sensors will undoubtedly improve performance, but at the cost of complexity and robustness. A more elegant solution will likely emerge from a deeper understanding of the fundamental mechanics of soft materials, and a willingness to embrace-rather than eliminate-the inherent uncertainties within these systems.

Original article: https://arxiv.org/pdf/2601.09104.pdf

Contact the author: https://www.linkedin.com/in/avetisyan/