SimTO-Robust: Automated Design of Soft Grippers Robust to Grasp Pose and Object Geometry

Kurt Enkera; Joshua Pinskier; Marcus Gallagher; David Howard

cs.ROOct 29, 2026

SimTO-Robust: Automated Design of Soft Grippers Robust to Grasp Pose and Object Geometry

Authors: Kurt Enkera, Joshua Pinskier, Marcus Gallagher, David Howard

Organizations: CSIRO Robotics, Data61, CSIRO, Australia. · School of Electrical Engineering and Computer Science (EECS), The University of Queensland, St Lucia, Australia.

Abstract

Robotic grippers operating in unstructured environments must robustly grasp objects of different shapes and sizes despite perception and positioning errors. Soft grippers that can achieve this robustness through intelligent mechanical design rather than sophisticated control software offer a promising solution. However, no existing computational design framework explicitly optimises soft gripper morphology for robustness across object geometries and grasp poses. We introduce SimTO-Robust, a generalisation of simulation-driven topology optimisation (SimTO) that simulates grasps across multiple objects and poses and aggregates their contact forces to guide gripper morphology optimisation. We generate 6,646 designs and physically evaluate 24 of them alongside 10 specialised SimTO designs, two rigid baseline grippers, and a baseline generalist soft gripper produced using conventional topology optimisation. All 37 grippers are tested in a pick-and-place task involving five feature-rich objects and six grasp poses, with external disturbances of increasing magnitude applied during the task. On average, SimTO-Robust designs complete pick-and-place more reliably than specialised SimTO designs and the rigid baselines. The best SimTO-Robust designs also withstood more of the applied disturbances than any other tested gripper. These results show that incorporating diverse grasp conditions into simulation-driven topology optimisation is a simple and effective approach to improving real-world soft gripper robustness. More broadly, they show that robust robotic grasping can be achieved through intelligent mechanical design rather than sophisticated sensing and control.

Figures & tables

Fig. 1 : Whereas SimTO [ 5 ] simulates a gripper grasping a single object in a single pose, SimTO-Robust (i) simulates grasps of $K$ objects in $P$ poses, (ii) combines the resulting contact forces, and (iii) uses them to update the gripper through topology optimisation. Iterating steps (i)-(iii) produces grippers with larger workspaces, grooved contact surfaces and other diverse features for more robust grasping than specialised SimTO designs and baseline rigid and soft grippers (Sec. IV ).

Fig. 2 : Numerical optimisation results. a) The three initial designs used in the SimTO-Robust experiments, and the reference generalist design from [ 11 ] . (b) The 12 SimTO-Robust experiments plus the SimTO baseline ( O1|P1|D1 ). Each marker is one generated design, and we plot its furthest simulated contact point (C.P.) in $x$ from the fixed port against the percentage of its C.P.s lying on $y=0$ . Outlined, enlarged markers are designs that were 3D-printed and validated in this work.

Design group	$N_{\text{d}}$	$O$	$Z_{1}$	$Z_{2}$	$X_{1}$	$Y_{1}$	$\theta_{1}$	Mean
O1\|P1\|D1	2071	92.3	77.2	43.3	76.7	36.6	75.8	67.0
O1\|P3\|D1	895	93.0	89.0	68.5	89.1	34.0	86.5	76.7
O1\|P5\|D1	1623	93.5	84.2	61.1	83.7	41.2	82.8	74.4
O1\|P1\|D1\|HL	2163	94.0	89.5	73.6	89.7	39.8	87.7	79.0
O5\|P1\|D1	346	92.7	83.1	61.7	83.3	44.8	82.7	74.7
O5\|P3\|D1	210	90.5	80.9	60.5	80.3	45.8	80.0	73.0

TABLE I : Simulated grasp robustness success (%) per test pose, averaged over all five objects and all five impulse directions.

N_{\text{d}}

: number of designs. Mean: average robustness over all six poses. The best value per column is emboldened.

Fig. 3 : Where contact occurs on the gripper shapes the optimised design. Simulated deformation of the initial and subsequent designs during optimisation of an hourglass specialist in (a) O1|P1|D1 and (b) O1|P1|D1|HL .

$P$	Designs	Curvy	Gear	Hourgl.	Spiky	Star	Mean
P1	Matched	60.4	79.0	39.1	57.9	83.3	63.9
	Other	64.5	75.0	60.9	56.5	77.7	66.9
	Gen.	73.4	80.6	66.8	70.4	82.5	74.7
P3	Matched	72.5	84.2	66.2	66.8	88.6	75.7
	Other	72.7	80.8	76.4	69.8	85.6	77.0
	Gen.	72.6	77.7	65.2	68.7	80.8	73.0

TABLE II : Simulated grasp robustness success (%) per test object, averaged over all poses and impulse directions, for the D1 groups in Table I . Matched : specialists (

K=1

) optimised for that object. Other : the four remaining object specialists in the group. Gen. : the generalist (

K=5

) with the same

P

. Mean: average robustness over all objects. For each

P

, the best value per column is emboldened.

Fig. 4 : Trade-off between grasp robustness and optimisation cost. Simulating more objects and poses increases the mean load-set size $\overline{N_{f}}$ and the computational cost. Grasp robustness, our measure of embodied intelligence, peaks at approximately 45 loads ( $K\times P=5$ ) and degrades beyond this point.

Fig. 5 : Experimental grasp robustness tests. (a): Soft grippers were mounted on a UR10e arm and actuated by a linear pneumatic actuator. (b): Our parallel-jaw rigid gripper. (c)-(e): In each trial, the object is first grasped at $t_{0}$ , transported during $t_{1}$ and placed at $t_{2}$ . It is then re-grasped and lifted during $t_{3}$ before undergoing four impulse-resistance tests, where the impulses grow in magnitude. First, the gripper is accelerated in all five directions shown in blue. After, the gripper is shaken at three increasing accelerations. If the object is still held, it is placed back on the target at $t_{4}$ . Bottom row: Representative failure modes: (f) the object sticking to gripper features; (g) asymmetric deformation at low $z$ grasp poses; (h) a robust design with a loose grip, causing failure during shaking; (i) the flat contact surface of the soft reference design fails to strongly grip objects; (j) a rigid gripper failing to place the object accurately in the $X_{1}$ pose.

Fig. 6 : Experimental grasping results. (a) All 37 validated grippers in their grasped state at pose $Z_{2}$ , grouped into baselines (yellow), SimTO specialists (pink), and SimTO-Robust designs (green). (b) Grasp score distribution of each design over its 30 grasps. Mean scores equal to or greater than that of the soft reference design are in bold.

Fig. 7 : Our most robust design. (a) Design 16 from Fig. 6 . (b) Fabricated design grasping the star.

Robotic Gripper Design

See all 14 papers

Sep 17, 2026

Graph-Based Design of Soft Grippers with Multi-Objective Quality-Diversity Optimisation

Effective manipulation across diverse objects is critical for applications ranging from agricultural harvesting to laboratory and domestic automation. While the inherent compliance of soft robotics is well suited to this challenge, designing grippers that generalize across tasks remains difficult due to the vast design space of continuum mechanics and the risk of overfitting to specific scenarios. We propose a graph-based design space for representing soft structures and mechanisms, coupled with a multi-objective, diversity-driven genetic optimization framework that explicitly promotes solution variety throughout the design process. Using multiple grasping scenarios during optimization, we study how task diversity influences the emergence of generalization to unseen objects and contact conditions. Our results show that optimization over a sufficiently diverse set of grasping cases leads to designs with emergent generalization, exhibiting improved robustness compared to task-specific solutions on novel scenarios. These findings suggest that diversity-driven optimization offers a principled pathway toward general-purpose soft grippers, aligned with the adaptable nature of soft robotics.

Andre Farinha, Ge Shi, Harry Bowman +3

CSIRO Robotics, Australia

Robotic Grasping Robotic Gripper Design Multi-Objective Evolutionary Optimization Generalization in Robotic Manipulation Quality-Diversity Optimization Soft Robotics

Ask PDF

Sep 3, 2026

Robot Aware Computational Design of Object Specific Passive Grippers for Additive Manufacturing

This paper presents an end-to-end computational pipeline that converts a selected object mesh, a measured object state, and a selected six-axis robot into an object-specific, unactuated, additively manufacturable gripper. The method couples exact-mesh RGB-D/ICP pose registration, deterministic surface-contact sampling, uncertainty-aware wrench screening, selection among six passive capture mechanisms, object-conformal surface synthesis, full-orientation robot inverse kinematics, a swept-volume-aware manufacturing domain, directional fused-deposition finite-element screening, and constrained three-dimensional SIMP topology optimization. Unlike workflows that treat grasp selection, tool geometry, motion, and structural design as separate problems, every exported design is bound to the source mesh, object pose, robot flange, contact set, and insertion hypothesis by a traceable design identifier. We derive the implemented registration, contact, fit-tolerance, finite-element, and density-optimization equations and prove three properties of the numerical construction: nodal load preservation, monotonic compliance sensitivity under SIMP interpolation, and voxel-domain containment after topology post-processing. Four archived object-specific attempts - a rabbit, camera flange, 3DBenchy, and faceted bust - meet the nominal fit, uncertain-wrench, runtime-sweep, and baseline/post-topology FEA gates. A deliberately enlarged +/-3 mm, +/-5 degree pose stress check differentiates the designs, retaining 29-134 of 160 simulated trials. Their reconstructed topologies retain 92.0-97.6% of the FE domain because functional regions are protected. Archived robot photographs show the corresponding printed assemblies qualitatively, while nominal material properties and absent coupon-calibrated, instrumented tests keep all four at digital-screening status rather than operational release.

Abdullah Yahya Abdullah Omaisan, Ibrahim Sheikh Mohamed

1Independent Researchers, Riyadh, Saudi Arabia. · 2QSS AI and Robotics Lab, Riyadh, Saudi Arabia.

Robotic Grasping Robotic Gripper Design Additive Manufacturing

Ask PDF

Aug 17, 2026

Design Optimization for Large High-Force Soft Robot Manipulators Under Gravitational Loads

Designing large soft robots capable of generating high forces for physical human-robot interaction remains a significant challenge in soft robotics. Prior work in large soft robots has focused on proof-of-concept prototypes, and no systematic framework exists for determining the suitability of a design paradigm for a desired task. This manuscript introduces a method for optimizing the geometry of a soft robot limb, maximizing its blocking force subject to an anti-bucking constraint under its own gravitational loading. We demonstrate that an explicit solution exists to the proposed optimization problem under certain assumptions. Experiments with three geometries of a large, soft, pneumatically-actuated manipulator demonstrate that the method correctly predicts which designs meet constraints and which produces the largest end-effector forces. This method, with its closed-form solution, can allow designers to determine a-priori if an intended class of soft manipulators is an appropriate choice for physical interaction at large size scales.

Isara Cholaseuk, Penelope Llibre, Alexa Kyriacou +5

Department of Mechanical Engineering, Boston University, Boston MA, USA · Division of Systems Engineering, Boston University, Boston MA, USA

Soft Robotics Constrained Optimization

Ask PDF

Abstract

Robotic Gripper Design

Explore similar work

Graph-Based Design of Soft Grippers with Multi-Objective Quality-Diversity Optimisation

Robot Aware Computational Design of Object Specific Passive Grippers for Additive Manufacturing

Design Optimization for Large High-Force Soft Robot Manipulators Under Gravitational Loads