In household manipulation, pose uncertainty often results in off-centre or partial initial contact, making reliable object acquisition difficult. Roller-based grippers can actively draw objects inward but often provide limited post-capture stability, whereas granular-jamming grippers require sufficient contact before jamming to achieve strong retention. This paper presents a hybrid roller-jamming gripper that integrates active object intake and post-capture retention within a single gripper. The proposed gripper uses inward roller rotation to increase contact and draw the object toward the gripper centre, followed by vacuum-induced granular jamming to stiffen the rollers and stabilise the grasp. The paper also presents a simplified geometric analysis of the gripper and a bench-level characterisation of the prototype's force capability. The gripper prototype was mounted on a 7-DoF robotic arm and evaluated using eight test objects. Furthermore, controlled planar position and orientation offsets were applied, with each condition repeated three times. The main evaluation comprised 840 grasp trials, including 216 planar-offset trials and 624 orientation-offset trials. Overall, the gripper succeeded in 812/840 trials: 215/216 planar-offset trials and 597/624 orientation-offset trials. The ablation evaluation comprised 162 trials on three objects. The roller-only and jamming-only conditions achieved 54/81 and 24/81 successes, respectively, showing their different contributions. These results provide initial mechanism-level evidence that hybrid roller-jamming is a promising strategy for improving acquisition and retention after imperfect first contact.
Figures & tables
System
Primary purpose
Acquisition mechanism
Retention mechanism
Actuation architecture
RBO Hand 2 [ 3 ]
Dexterous grasping through compliance and underactuation
Pneumatic bending of compliant fingers and palm around the object
Compliant enveloping contact maintained by pneumatic pressure
Seven pneumatic actuators grouped into four control channels
Two-fingered hand with jamming finger tips [ 10 ]
Adaptive grasping with variable-stiffness fingertips
Finger closure with passive fingertip conformity
Vacuum-jammed fingertips
Four servomotors and a pneumatic jamming system
Simple Belted Parallel-Jaw Gripper [ 16 ]
Grasp-to-grasp in-hand translation and reorientation
Base parallel-jaw closure
Force-controlled jaw clamping with belt and finger-surface contact
Four independent belt motors plus the base-gripper actuator
Roller Grasper V3 [ 19 ]
Full 6-DoF in-hand manipulation
Finger closure followed by rolling pickup
Four-finger rolling contact, assisted by an active palm
Four fingers with three active DoF each; 12 active DoF
Mecanum-wheel hand [ 20 ]
Omnidirectional in-hand manipulation without finger gaiting
Coupled four-finger opening and closing
Force closure with spring-loaded wheel constraint forces
Three-servo coupled reconfiguration and two wheel-drive motors
Proposed roller-jamming gripper
Grasp acquisition and retention under pose offsets
Shared finger closure followed by inward rolling intake
Vacuum-induced granular jamming of the same rollers
Three active DoF and pneumatic vacuum switching
TABLE I: Comparison with representative compliant, jamming-based, and active-surface grippers. The systems address different tasks and are compared primarily by their mechanisms and actuation requirements.
Fig. 1: Prototype of the proposed roller-jamming gripper mounted on a 7-DoF robot arm (xArm7).
Fig. 2: Mechanical architecture of the proposed gripper: (Left) overall assembly and (Right) base module, showing the gripper palm, mounting adapter, XM430 actuator, and worm-gear finger-closing transmission.
Fig. 3: Roller-finger architecture: (Left) roller-finger module with the XL320-driven worm-gear transmission and (Right) spherical membrane roller construction, including the clamp rings, drive interface, and rotary union for vacuum feedthrough.
Item
Value
Overall envelope (W × L × D)
70×228×177mm
Total gripper mass
1.31kg
Spherical membrane roller diameter
≈80mm
Roller axial width
73mm
Finger pivot to roller centre
115.6mm
Max roller centre-to-centre distance
285mm
TABLE II: Key mechanical specifications of the gripper.
Fig. 4: Membrane mould components: (Left) internal core and alignment shaft and (Right) assembled two-part mould containing the core shell and outer cavity used to form the spherical dimpled membrane.
Fig. 5: Control, communication, electrical-power, and pneumatic architecture of the prototype gripper system.
Fig. 6: Simplified front-view rigid-envelope model of the gripper. The fingers rotate symmetrically about pivots separated by dp , with pivot-to-roller-centre distance L , roller radius R , and roller-centre spacing dc .
Fig. 7: Nominal fully closed geometry used to estimate the minimum supported circular-feature radius rmin . The undeformed roller envelopes have radius R and centre-to-centre spacing dc(θmin) .
Fig. 8: Simplified front-view tilted-contact geometry used to estimate the minimum supported circular-feature radius. The roller pair is tilted by Δθ while maintaining the nominal centre spacing dc(θmin) . The supported feature remains tangent to both undeformed roller envelopes and the support plane.
Fig. 9: Preliminary bench force tests. Force time histories from the single roller-traction, grasp slip-out-resistance, and clamping-force trials. The maximum measured forces were 8.4, 25.6, and 27.7 N, respectively.
Fig. 10: Finite-state machine of the grasp controller. CLOSE and HOLD use motor-current and dwell conditions, whereas INTAKE and JAM use prescribed durations. Abort or manual-release commands return the system to RELEASE .
Fig. 11: Experimental setup (top) and example of camera-based object localisation (bottom). For the T-shirt, pressure gauge, and cable, the overhead camera estimates the object centroid on the tabletop and in-plane extents after each manual reset. AprilTags define the tabletop reference frame used for localisation and offset generation.
Fig. 12: Test objects used for evaluation: pen, marble, gallon jug, plastic cup, paper sheet, T-shirt, pressure gauge, and cable
Object
Mass [g]
Physical characteristics
Contact challenge represented
Pen
24.6
Rigid, slender, and low-profile
Shallow contact with a narrow object.
Marble
19.9
Rigid, smooth, small, and spherical
Localised contact and enclosure of a small sphere.
Gallon jug
420
Rigid, large, and relatively heavy
Retention of a large object under higher loading.
Plastic cup
6.0
Lightweight, thin-walled, and deformable
Grasping without excessive deformation.
Paper sheet
4.2
Thin, flat, flexible, and low-profile
Shallow and potentially one-sided contact.
T-shirt
180
Highly deformable textile
Changing shape and contact region during acquisition.
TABLE III: Physical characteristics and contact challenges represented by the test objects.
Object
Planar/ablation grasp height [mm]
Orientation grasp height [mm]
Intake duration [s]
CLOSE current threshold [mA]
HOLD current threshold [mA]
Pen
6
6
5
40
200
Marble
10
10
5
130
230
Gallon jug
254
254
2
80
230
Plastic cup
58
58
4
40
200
Paper sheet
-10
-10
5
70
70
T-shirt
0
6
8
40
200
TABLE IV: Object-specific parameters used in the repeated experiments.
Test protocol
Successful trials
Success rate
95% CI
Planar offset
215/216
99.5%
97.4–99.9%
Orientation offset
597/624
95.7%
93.8–97.0%
Main evaluation
812/840
96.7%
95.2–97.7%
JAM OFF
54/81
66.7%
55.9–76.0%
ROLLER OFF
24/81
29.6%
20.8–40.3%
TABLE V: Success rate across the test protocols with two-sided 95% Wilson score confidence intervals.
Fig. 13: Successful grasps with the roller-jamming gripper on the eight test objects used in evaluation. Top row (left to right): pen, marble, gallon jug, plastic cup. Bottom row (left to right): paper sheet, T-shirt, pressure gauge, cable.
Fig. 14: Planar-offset test results for the eight objects, with three trials per condition. Each coloured marker corresponds to one offset condition repeated three times; the marker colour indicates the number of successful trials out of three. Offsets are applied on a 3×3 grid about the estimated object centre and scaled to ±20% of the corresponding in-plane object extents.
Fig. 15: Ablation results for the pen, gallon jug, and cable under the planar-offset protocol. The top row shows the roller-only condition (JAM OFF), and the bottom row shows the jamming-only condition (ROLLER OFF). Marker colour indicates the number of successful trials out of three at each offset.
Fig. 16: Orientation-offset test results across the roll, pitch, and yaw conditions, with three trials per condition. Each cell represents one orientation-offset condition repeated three times, and the number and colour in each cell indicate the number of successful trials out of three. The 3×3×3 roll–pitch–yaw grid excludes the all-zero condition, shown in grey as not tested. For the paper sheet, the orientation-offset magnitude was reduced to ±5∘ to avoid gripper–table collision.
Robots are increasingly used in unstructured environments. The need for them to safely grasp unknown objects without damaging them becomes crucial. Humans achieve this by sensing and quickly responding by adjusting their grasping force. Similarly, effective grasp acquisition in robots requires compliant interaction strategies that can adapt to uncertain object properties and adjust to any instabilities during manipulation. We present a geometry-aware force/torque-based contact estimation method for a coarse position-controlled gripper, combined with an adaptive admittance controller for safe grasp acquisition. The desired contact forces are estimated online to keep stable contact with objects of unknown properties. This enables compliant and stable grasps while avoiding excessive forces. Experiments with objects of different sizes, shapes, stiffnesses, and weights show that the proposed algorithm not only prevents slippage but also applies minimal force to safely grasp an object without causing excessive deformation.
Marko Pavlic, Moritz Geier, Timo Markert +1
Machine Vision and Perception Group, Chair of Robotics, Artificial Intelligence and Real-time Systems, Technical University of Munich, Munich, Germany · Resense GmbH, Klingenberg, Germany · Semantic Information Systems Group, Osnabrück University, Osnabrück, Germany
Continuum manipulators provide dexterous motion in confined spaces, but structural compliance, hysteresis, and load-dependent deformation leave residual position and orientation errors that can undermine reliable contact with rigid grippers. To address this limitation, this paper presents a lightweight support-enhanced granular-jamming gripper tailored to a continuum manipulator. The gripper maintains compliance before jamming while establishing a direct load path to the continuum manipulator tip after jamming. To improve its grasping performance, we systematically designed membrane materials, particles, filling ratios, and the internal support structure, and further identify geometry-dependent grasp boundaries with respect to contact offset and object shape. Building on these results, we construct a physical manipulation system integrating the continuum manipulator, granular-jamming gripper, visual feedback, tendon actuation, and pneumatic control. We then train a reinforcement-learning-based reaching controller in a randomized simulation and deploy it on the physical system, demonstrating how positioning control and contact level mechanical adaptation can complement each other in a modular grasp-and-release task. By introducing an adaptive structure that relaxes the need for highly accurate modeling and positioning control, this work explores a design paradigm that integrates physical and embodied intelligence.
Controlling friction at the fingertip is fundamental to dexterous manipulation, yet remains difficult to realize in robotic hands. We present the design and analysis of a robotic fingertip equipped with passive rollers that can be selectively braked or pivoted to modulate contact friction and constraint directions. When unbraked, the rollers permit unconstrained sliding of the contact point along the rolling direction; when braked, they resist motion like a conventional fingertip. The rollers are mounted on a pivoting mechanism, allowing reorientation of the constraint frame to accommodate different manipulation tasks. We develop a constraint-based model of the fingertip integrated into a parallel-jaw gripper and analyze its ability to support diverse manipulation strategies. Experiments show that the proposed design enables a wide range of dexterous actions that are conventionally challenging for robotic grippers, including sliding and pivoting within the grasp, robust adaptation to uncertain contacts, multi-object or multi-part manipulation, and interactions requiring asymmetric friction across fingers. These results demonstrate the versatility of passive roller fingertips as a low-complexity, mechanically efficient approach to friction modulation, advancing the development of more adaptable and robust robotic manipulation.
Ethan Fisk, Taeyoon Lee, Shenli Yuan
Department of Mechanical Engineering, Northeastern University, Boston, MA 02115, USA · RAI Institute, Cambridge, MA 02142, USA