Applying adhesive tape to secure wire harnesses or seal packages requires robots to coordinate a flexible strip, a moving roll, and surfaces that attach and detach. Simulation could make these interactions repeatable for robot development and evaluation, but resolving every adhesive layer is expensive and can suppress roll motion at practical solver tolerances, while a permanently rigid roll cannot release material. We present TapeSim, a tape simulator that concentrates deformation near the unwinding region and along the released strip. We will release the source code. A rigid cluster represents most wound material, while an advancing deformable collar enables payout and leaves released tape flexible and reattachable. Optional releasable bonds simplify adhesive interfaces and reduce mean step times for smaller rolls. Controlled swing tests show improved roll rotation. At 32 turns, clustering gives 3.2-3.4x mean physics-step speedups at a fixed Newton tolerance and 4.5-8.4x for comparable roll motion. Across five real-motion Stick replays, the clustered variants reduce mean image-plane core-landmark error by 23-29% relative to the full-shell cohesive baseline. On 100 paired Peel cases, they improve balanced accuracy from 50% to 72.9-76.3%, with interface rankings varying across tasks. A teleoperated box-sealing sequence demonstrates attachment, dispensing, cutting, and sealing in a continuous workflow.
Figures & tables
Fig. 2: Peel-force calibration and tape-handling feasibility. (a) Force-gauge measurement and simulated peel shape at 50 s; the simulated fixture gives a median reaction of 2.54 N at Cn=Ct=0.675 . The arrow indicates pulling direction; the simulated profile uses equal horizontal and vertical scales. (b) A 32-turn Bond-R roll dispenses tape as the collar advances. Orange/magenta/blue indicate cluster/collar/free end, and gray the hub. State bars span the same rest-material interval; black markers track one material row. Roll views share a common scale. (c) Blade indentation and separation of a fixed-end strip. Black edges are fixed and gray denotes the blade. Color denotes positive maximum principal membrane tension (N/m), showing tension release after contact-pressure-triggered cutting.
Quantity
Value
Source
Tape width
19 mm
Measurement
Full film thickness
0.18 mm
Measurement
Core inner / outer diam.
38.0 / 42.2 mm
Measurement
Core axial height
20.0 mm
Measurement
Tape density
1300 kg/m 3
Measurement
Membrane / bending E
50 / 50 MPa
Configured
TABLE I: Parameter settings for the 19-mm tape assets.
Fig. 3: Roll motion and computational cost. (a) Mean physics-step time over 1750 steps at Newton tolerance 10−3 m/s, excluding cluster updates and rendering. (b) Mean physics-step time for comparable roll motion, using tolerances selected by matching hub-rotation ranges. (c) Hub rotation after release for a 32-turn roll at 10−2 m/s, relative to the first post-release state. (d) Coh-F at 7.48 and 12 s, before and after hub release at 7.5 s. Views share time and scale; dots mark the held tip, circles the hub center, and spokes the hub orientation.
Fig. 4: Clustered models retain Stick replay fidelity. Top: ego-camera core-landmark error for five real-motion replays; each curve averages two simulation repeats. Time is measured from recording start, excluding the stationary first 3 s. Hollow markers identify estimated real landmarks. Bottom: synchronized real/sim frames from Stick 19, repeat 1, at 6.5 and 9.5 s, using the same crop across models.
Fig. 5: Real versus simulated policy success across three tape tasks. Top: Coh-R rollout scenes. Bottom: simulated success rate (horizontal) versus real success rate (vertical); the dashed diagonal indicates equal rates. Circles/squares denote Peel/Stick, and filled/open triangles denote Wrap success with at least one/two completed turns. Peel, Stick, and Wrap use 100, 71, and 55 matched cases, respectively.
Fig. 8: Continuous teleoperated box sealing. Frames from one simulated rollout show flap closing, tape attachment, dispensing, cutting, and pressing the seal. Times are measured from the start of the recording.
Simulator-in-the-loop optimization offers a promising inference-time mechanism for robot manipulation. It uses a physical simulator as a backend rollout engine to evaluate candidate trajectories in parallel and refine nominal actions online, a paradigm shown to be effective in rigid-body manipulation where state and contact are relatively tractable. We bring this paradigm to real-world cloth manipulation from a single RGB input through three pillars. (i) We design a scalable synthetic-data generation and inference-time rollout pipeline built on FLASH, a deformable-object simulator that provides a practical balance among physical fidelity, numerical stability, and rollout efficiency. (ii) We develop a real-to-sim module, trained purely on synthetic data, that maps a single RGB observation to simulation-compatible cloth state by fusing pretrained visual features with learnable canonical tokens. (iii) We perform online planning by coupling a sparse-mesh rollout backend with prior-guided MPPI, anchored at an offline-distilled policy trajectory, preserving manipulation-relevant deformation and contact while enabling sufficient parallel rollout batches. Real-robot experiments show higher success rates than baseline methods and closed-loop correction under mid-fold perturbations. Project page: https://silr-cloth.github.io/
Xin Liu, Yulin Li, Ziming Li +7
National University of Singapore · Shanghai Jiao Tong University
Contact-rich manipulation requires robots to regulate both motion and interaction forces, yet achieving adaptive compliance remains a fundamental challenge. Learning from real-world data is costly and risky, while simulation-based approaches struggle with the sim-to-real gap in contact dynamics; existing sim-to-real methods either require real-world adaptation or sacrifice adaptive compliance by relying on isotropic compliant controllers. Our key insight is that force regulation decomposes into a time-varying but simulation-transferable directional component and a dynamics-sensitive but manually tunable magnitude component. We instantiate this directional component as two policy outputs, a task frame and a control mode vector, predicted by a visuomotor policy adapted from a pre-trained VLA model and trained via imitation learning on automatically generated simulation demonstrations. At deployment, an admittance controller integrates these predictions with human-specified stiffness and target wrench values to realize adaptive compliance. Our approach achieves adaptive compliance using only simulation data and can benefit from large-scale VLA pre-training. Extensive real-world experiments on four contact-rich tasks, microwave opening, peg-in-hole insertion, whiteboard wiping, and door opening, demonstrate strong task success rates and robustness to external disturbances. Project page: https://yifei-y.github.io/project-pages/TDC/.
Yifei Yang, Anzhe Chen, Zhenjie Zhu +6
Zhejiang University · Zhejiang Humanoid Robot Innovation Center
RGB sim-to-real for deformable manipulation has remained largely unsolved without real-world fine-tuning. We present SimWeaver, which trains zero-shot RGB VLA policies on 200 simulated demonstrations per task, reaching above 80% per-task and 91% average real-world success across 5 diverse deformable tasks including plastic-bag manipulation, without teleoperation or per-task calibration. SimWeaver combines a reliable measurement-backed simulator (SimWeaver-Sim) with an extensible asset framework supporting single-image generation(SimWeaver-Asset), a deterministic topology-aware trajectory synthesizer (SimWeaver-Syn), and a sim-to-real protocol with ISP-aware photometric augmentation (SimWeaver-Real). On silk grasping, the sim-trained policy reaches 100% under visual distribution shifts where real-data baselines drop to 9-70%, at two orders of magnitude lower per-trajectory cost. We will release SimWeaver and a representative asset subset. Project page: https://simweaver.github.io/
Wenkang Hu, Haoran Wang, Yitong Li +10
1Shanghai Jiao Tong University · 2Horizon Robotics · 3Style3D Research