(last updated March 11th, 2026)
A widely used perspective in the literature classifies deformable objects based on their physical structure and manipulation characteristics. In the survey by José Sanchez and colleagues, deformable objects are categorized into several types such as linear objects (e.g., cables, ropes, and wires), planar or sheet-like objects (e.g., cloth, fabrics, and thin materials), and three-dimensional deformable objects including soft volumetric items like food products, foams, or biological tissues.
Each category presents different manipulation challenges and typically requires distinct sensing, modeling, and control strategies. Linear objects often involve tasks such as routing or knotting, sheet-like objects require operations such as folding or unfolding, while volumetric soft objects demand careful force control to avoid excessive deformation or damage. Across these categories, effective manipulation relies on accurate perception of object state, deformation modeling, and the integration of multimodal sensing—including vision, force, and tactile feedback—to guide interaction with the material.
Manipulating deformable objects remains one of the most challenging problems in robotics. Unlike rigid objects, deformable items such as cables, textiles, food products, and biological tissues change shape during contact. Their configuration cannot be described by a small number of parameters, which makes perception, modeling, and control significantly harder. Recent surveys emphasize that deformable object manipulation (DOM) requires coordinated advances in hardware design, sensing, deformation modeling, planning, and learning-based control
Among these components, the end-effector hardware—grippers, robotic hands, and specialized tools—plays a central role, since it forms the physical interface between the robot and the deformable material.
The importance of specialized end-effectors
Traditional robotic manipulation has largely focused on rigid objects, where standard parallel grippers are often sufficient. In contrast, deformable objects typically require task-specific or custom-designed grippers because their manipulation involves actions such as stretching, folding, sliding, or shaping. Examples include grippers that allow cables to slide, flat clips designed for fabrics, or soft tools for biological tissues.
This diversity arises because deformable objects behave very differently depending on their physical properties. Materials may be elastic, plastic, compressible, or highly flexible, meaning that the same grasping strategy cannot work across all domains. As a result, robotic systems often combine specialized hardware with sensing and adaptive control (reference).
Classical grippers for deformable objects
The simplest solutions rely on adaptations of standard industrial grippers. Parallel pinch grippers remain widely used due to their robustness and ease of control. They are effective for basic pick-and-place tasks involving cloth pieces, cables, or soft packaging. However, these grippers provide limited control over local deformation and may slip or damage delicate materials.
Vacuum suction grippers are another common option, particularly for sheet-like objects such as textiles or thin plastic layers. Suction tools allow fast and reliable grasping of flat surfaces, but they are sensitive to surface porosity, wrinkles, and irregular shapes.
In some applications, roller-based feeding mechanisms or sliding tools are used to guide materials such as fabrics or wires. These tools support continuous manipulation rather than simple grasping, enabling actions like alignment, feeding, or controlled stretching. However, these systems tend to be highly specialized and difficult to generalize across tasks.
Anthropomorphic robotic hands
A promising research direction is the development of anthropomorphic robotic hands with multiple fingers and higher dexterity. Multi-fingered hands allow robots to perform complex manipulation actions such as regrasping, in-hand repositioning, and shape control. These capabilities are particularly important for deformable objects, where manipulation often involves controlling how the object bends or stretches rather than simply lifting it.
Despite their potential, anthropomorphic hands still face several limitations in deformable manipulation. Many existing designs were originally developed for rigid-object grasping and lack sufficient compliance or tactile sensing. As a result, controlling contact forces precisely can be difficult, and manipulation strategies often require complex perception and control algorithms.
Recent research therefore emphasizes integrating tactile sensing, compliant mechanisms, and adaptive control into robotic hands. The goal is to approximate the versatility of the human hand, which remains the most capable tool for handling deformable materials.
Soft robotic grippers
Soft robotic grippers represent another important hardware approach. These devices use compliant materials such as silicone or flexible polymers to create fingers that naturally conform to an object’s shape. Because the contact surface adapts to the geometry of the object, soft grippers distribute pressure more evenly and reduce the risk of damage.
Common actuation methods for soft grippers include pneumatic chambers, tendon-driven compliant structures, and granular jamming mechanisms. These technologies allow the gripper to switch between flexible and rigid states or to adapt its shape dynamically during grasping (reference).
Soft grippers are particularly effective for handling fragile items such as fruit, food products, or biological tissues. However, they generally offer less precision and lower gripping force than rigid grippers, which can limit their usefulness for tasks that require exact positioning or strong manipulation forces.
Sensorized grippers and tactile hardware
A key trend identified in recent research is the integration of sensing directly into the gripper hardware. Because deformable objects change shape during manipulation, visual perception alone is often insufficient. Cameras may be occluded by the robot’s fingers or unable to capture small local deformations.
To address this problem, modern grippers increasingly incorporate tactile sensors, force–torque sensors, slip detection systems, and proprioceptive feedback. These sensors allow the robot to measure contact forces, detect object motion within the grasp, and adjust its grip in real time.
Such multimodal sensing is essential for reliable manipulation of deformable materials. According to recent surveys, combining tactile sensing with learning-based methods is one of the most promising directions for improving robotic dexterity in deformable manipulation tasks.
Gripper for textile objects
A more focused perspective on end-effector design for deformable objects can be found in works specifically addressing textile manipulation. Early studies, such as the review by Koustoumpardis and Aspragathos (2004), provide a systematic analysis of gripping devices for fabric handling, particularly in industrial contexts. This work categorizes a range of mechanisms—including clamping, suction, needle-based, and airflow-based approaches—and highlights the strong dependence of gripper performance on fabric properties such as thickness, porosity, and friction. Borràs et al. (2020), that offers a more recent bibliographic review focused on robotic grasping of textiles, emphasizing the diversity of gripper designs proposed in the robotics literature. These include adaptations of pinch and suction grippers as well as more specialized devices for edge grasping, sliding, and controlled interaction with deformable surfaces. Together, these works underline that textile manipulation requires careful consideration of both global shape control and local contact mechanics, reinforcing the broader observation that gripper design remains highly task- and material-dependent.
Outlook
Despite significant progress, deformable object manipulation remains an open challenge in robotics. Current systems often rely on specialized grippers tailored to specific materials or tasks. Achieving a more general solution will likely require combining compliant hardware, rich tactile sensing, and advanced learning-based control strategies.
Future robotic end-effectors will therefore evolve from simple mechanical tools into sensor-rich, adaptive interfaces capable of interacting safely and intelligently with complex deformable materials. Advances in these technologies will be essential for enabling robots to operate effectively in real-world environments such as manufacturing, healthcare, and service robotics.
References
Sanchez, J., Corrales, J. A., Bouzgarrou, B.-C., & Mezouar, Y. (2018). Robotic manipulation and sensing of deformable objects in domestic and industrial applications: A survey. The International Journal of Robotics Research, 37(7), 688–716. https://doi.org/10.1177/0278364918779698
Gu, F., Zhou, Y., Wang, Z., Jiang, S., & He, B. (2023). A survey on robotic manipulation of deformable objects: Recent advances, open challenges and new frontiers. arXiv. https://doi.org/10.48550/arXiv.2312.10419
Zhu, J., Cherubini, A., Dune, C., Navarro-Alarcon, D., Alambeigi, F., Berenson, D., Ficuciello, F., Harada, K., Li, X., Pan, J., & Yuan, W. (2022). Challenges and outlook in robotic manipulation of deformable objects. IEEE Robotics & Automation Magazine. https://doi.org/10.1109/MRA.2022.3147415
Arriola-Rios, V. E., Guler, P., Ficuciello, F., Kragic, D., Siciliano, B., & Wyatt, J. L. (2020). Modeling of deformable objects for robotic manipulation: A tutorial and review. Frontiers in Robotics and AI. https://doi.org/10.3389/frobt.2020.00082
Shintake, J., Cacucciolo, V., Floreano, D., & Shea, H. (2018). Soft robotic grippers. Advanced Materials, 30(29). https://doi.org/10.1002/adma.201707035
Zaidi, S., Maselli, M., Laschi, C., & Cianchetti, M. (2021). Actuation technologies for soft robot grippers and manipulators: A review. Current Robotics Reports. https://doi.org/10.1007/s43154-021-00054-5
Koustoumpardis, P. and Aspragathos, N. (2004), A review of gripping devices for fabric handling. Proc. Int. Conf. Intel. Manipulation Grasping, pp. 229–234. (here)
J . Borràs, J., Alenyà, G. and Torras, C. (2020) A grasping-centered analysis for cloth manipulation. IEEE Transactions on Robotics, 36(3): 924-936.
https://doi.org/10.1109/TRO.2020.2986921