The case for flexibility is rooted in safety and adaptability. Rigid grippers, built from materials like steel and aluminum, excel at exerting high force through inflexible joints and links, but their lack of compliance makes them poorly suited to delicate or irregularly shaped objects—and potentially hazardous around humans. Soft grippers, by contrast, are constructed from pliable materials such as silicone, elastomers, and plastics, allowing them to deform around whatever they grasp. This conformability provides a secure yet gentle hold without damaging the object, which is precisely why the review anticipates widespread adoption in sectors where safety is paramount, including healthcare and any application involving direct human-robot interaction. The authors define flexible grippers broadly as systems capable of shape adaptation, whether through soft materials, compliant structural designs, or hybrid combinations of both.

Among the most striking designs catalogued is the grain-filled flexible ball gripper, often called a universal gripper. It consists of a soft balloon packed with granular material—coffee grains in one notable implementation—that passively conforms to an object’s contour when compressed, then hardens under vacuum assistance to lock the shape in place. Releasing requires only positive pressure, which softens the membrane and loosens the grasp. The elegance of this design lies in its simplicity: no motors, cables, or linkages, just a single pump regulating airflow. Its limitation is equally clear—proficiency with lightweight objects under one kilogram, but not the optimal choice for heavier items where mass matters.

Hybrid grippers attempt to resolve the fundamental trade-off between softness and strength. Soft grippers offer a larger contact area and greater range of motion than rigid counterparts, but they suffer from reduced fingertip forces and slower actuation speeds. By embedding rigid materials into soft structures, researchers have dramatically closed this gap. In comparative testing with pneumatic actuation pressures ranging from 0 to 80 kilopascals, a hybrid design exhibited fingertip pressure 1.5 to 2 times greater than alternative soft gripper models, including Hybrid PneuNet, fast-PneuNet, and fiber-reinforced variants. The hybrid gripper successfully lifted a basket weighing up to 1.25 kilograms, demonstrating that structural reinforcement need not sacrifice the adaptability that makes soft systems attractive in the first place.

Electro-adhesive grippers represent perhaps the most counterintuitive approach: they grip without pressure at all. These devices use flexible adhesive pads made from polyurethane, silicone, or polyimide, with embedded electrodes that generate an electric field when a high voltage—typically in the kilovolt range—is applied. The charged pad creates electrostatic adhesion that secures objects against thin or flat surfaces. Remarkably, despite the high voltage, energy consumption remains below one watt because only a very low current passes through the pad. The electrodes themselves are thin and flexible, preserving the gripper’s overall compliance. The review identifies this technology as ideal for fragile, lightweight items, though performance can degrade on rough surfaces or in dusty, oily environments, and a continuous power supply is required to maintain the grip.

Magnetic approaches offer yet another pathway to compliant grasping. One design consists of an elastic membrane filled with water, with a permanent magnet at the bottom and an electromagnet at the top. When pressed against an object, the membrane conforms to its shape while the magnets attract each other, securing the hold; release is achieved by energizing the electromagnet to generate a repulsive field. Energy is consumed only during release, and experimental tests showed that fragile objects such as plastic bottles and quail eggs could be gripped safely without damage. A related approach integrates magnetorheological fluid—a smart material that behaves like a liquid under normal conditions but rapidly solidifies in a magnetic field—into the gripper membrane. This allows the gripper to adjust its stiffness on demand, with tests showing gripping force increasing from 3.8 to 7.1 newtons, enabling safe manipulation of both fragile and relatively heavy objects.

How these grippers are powered matters as much as how they are shaped. The review categorizes actuation into four primary types, each with distinct trade-offs. Vacuum grippers, relying on suction cups made from polyurethane, rubber, or foam, handle payloads from 0.1 to 20 kilograms and excel on flat or slightly curved surfaces like glass and marble, making them common in food processing where contamination avoidance is critical. Hydraulic systems, using incompressible fluids like oil or water, deliver the highest force output—handling payloads up to 11.3 kilograms—but require regular maintenance and careful design to prevent electrical leakage when water is the working fluid. Pneumatic grippers, which use air pressure within elastomeric chambers to bend fingers, offer low-pressure operation and adaptability across payloads from 0.1 to 9 kilograms, though they demand a continuous compressed air supply. Servo-electric grippers, controlled by microprocessors, provide the highest precision and fastest response times but generate the lowest gripping forces, suited to light and medium tasks.

Sensor integration transforms these mechanical systems into intelligent ones. Force/torque sensors, typically positioned between the robot’s wrist and the gripper, measure interaction forces across six degrees of freedom using piezoresistive, piezoelectric, or capacitive technologies, enabling real-time feedback that allows dynamic and compliant responses to unplanned physical contact. Pressure sensors regulate the contact between object and gripper, with load cells, pressure-indicating film, and tactile arrays each offering different approaches to quantifying force per unit area. RGB-D sensors capture both color imagery and depth information, allowing the robot to localize objects in three-dimensional space and verify that the gripper has made proper contact. Perhaps most intriguingly, sensorless sensing techniques detect object interactions by measuring internal system parameters—motor current, voltage variations, torque, vibration—eliminating the need for external sensors entirely. One validated approach achieved contact detection with 0.2 newton sensitivity in a hydraulic gripper using only a disturbance observer analyzing internal pressure deviations.

Control strategies span from classical to cutting-edge. PID controllers, prevalent in contemporary industry, use proportional, integral, and derivative feedback to minimize error between setpoint and actual state, remaining reliable in structured environments where system dynamics are well-defined. Fuzzy logic extends beyond binary true-false paradigms by incorporating intermediate values, offering a robust framework for handling uncertainty—particularly useful when grasping deformable or irregularly shaped objects where traditional mathematical modelling is challenging. Machine learning approaches, including convolutional neural networks for object recognition and reinforcement learning for adaptive grasping, enable robots to autonomously refine their strategies based on feedback. Deep reinforcement learning architectures such as the Grasp-Q-Network process data from multiple cameras and sensors to determine optimal grasping strategies, with reinforcement learning-based approaches achieving a 98.8 percent success rate in regrasping tasks and trajectory prediction combined with reinforcement learning increasing dynamic object grasping success rates up to 83 percent.

The review’s authors propose a systematic design workflow spanning project requirements, concept development, control strategy formulation, simulation, prototyping, and iterative testing—a structured process ensuring each gripper is tailored to specific operational demands. Looking forward, they identify key challenges in optimizing payload capacity, improving energy efficiency, and enhancing adaptability for complex tasks. Future research will focus on electroadhesive grippers and 3D-printed flexible finger-type designs, with efforts directed toward enhancing adhesion force, surface adaptability, and power efficiency while addressing high-voltage requirements. The trajectory is clear: as robots increasingly share spaces with humans and handle objects of unpredictable shape and fragility, the gripper—the humble interface between machine and world—may prove to be the most important innovation of all.

Subject of Research: A review of flexible robotic gripping systems

Article Title: A review of flexible robotic gripping systems

Article References: Tarhan, B., Asif, S., Webb, P., & Chacin, M. (2025). A review of flexible robotic gripping systems. Frontiers in Manufacturing Technology, 5, Article 1517727. https://doi.org/10.3389/fmtec.2025.1517727

Image Credits: AI Generated

DOI: 10.3389/fmtec.2025.1517727

Keywords: review, flexible, robotic, gripping, systems, scientific research