DexArena: A Unified Benchmark for Long-Horizon Dexterous Manipulation
Organizations: University of Michigan – Ann Arbor · Elastic Robotics · Shanghai Institute for Physical Artificial Intelligence and Robotics (PAIR) · National University of Singapore ·
Abstract
Long-horizon dexterous manipulation remains a fundamental challenge for robot learning, requiring the composition of multiple skills, precise bimanual coordination, generalizable visual understanding, and robust skill transitions. Yet existing manipulation benchmarks largely emphasize isolated skills or support only a limited range of end-effectors, making these capabilities difficult to evaluate systematically. We introduce DexArena, a benchmark designed to evaluate learning and generalization in long-horizon bimanual dexterous manipulation. DexArena features 10 long-horizon tasks and 30 stage-level tasks spanning diverse household and industrial scenarios. Each long-horizon task is composed of explicitly defined stages with physically grounded success conditions, enabling reliable evaluation of both intermediate progress and end-to-end task completion. A unified benchmark interface and verified human-teleoperated demonstrations enable testing different policy learning algorithms, with standardized simulation configurations, scene randomization, trajectory replay, and task-level evaluation. Through extensive baseline experiments, we identify substantial challenges in precise manipulation, generalization to scene variation, and skill composition. DexArena establishes a challenging and reproducible benchmark for developing robot learning methods that move toward generalizable long-horizon dexterous behavior.
Figures & tables
| Benchmark | Num. of tasks | Dexterous hands | Bimanual tasks | Cross- embodiment | Teleop demos | Long- horizon | Task composition |
|---|---|---|---|---|---|---|---|
| Adroit [ 8 ] | 4 | 100 | |||||
| RLBench [ 3 ] | 100 | ||||||
| Meta-World [ 4 ] | 50 | ||||||
| ManiSkill2 [ 5 ] | 20 | ||||||
| DexArt [ 9 ] | 4 | ||||||
| RoboCasa [ 6 ] | 100 |
| Long-horizon | Stages | Demo-guided RL | Imitation learning | VLA models | |||
|---|---|---|---|---|---|---|---|
| PPO | RLPD | ACT | DP | GR00T | (LoRA) | ||
| Meal preparation | OpenLunchBox | 74 | 100 | 100 | 96 | 100 | 100 |
| PlaceBread | 12 | 86 | 88 | 20 | 60 | 70 | |
| ScoopTomato | 0 | 0 | 44 | 0 | 12 | 32 | |
| Cabinet and pouring | OpenCabinet | 62 | 100 | 94 | 88 | 92 | 98 |
| PickBottle | 0 | 14 | 68 | 18 | 78 | 54 | |
| Long-horizon | Stages | Demo-guided RL | Imitation learning | VLA models | |||
|---|---|---|---|---|---|---|---|
| PPO | RLPD | ACT | DP | GR00T | (LoRA) | ||
| Meal preparation | OpenLunchBox | ||||||
| PlaceBread | |||||||
| ScoopTomato | |||||||
| Cabinet and pouring | OpenCabinet | ||||||
| PickBottle | |||||||
Appendix figures & tables15 assets
Supplementary material from the paper’s appendix.
Appendix
| Category | # | Stages | Time (s) | Contact ratio | Fingers | |
|---|---|---|---|---|---|---|
| Opening and closing | 7 | OpenCabinet | OpenDrawer | 3.0–8.1 | 0.52 | 2.3 |
| OpenFridge | CloseFridge | |||||
| OpenLunchBox | OpenLaptop | |||||
| OpenPackage | ||||||
| Grasping and placement | 8 | PickBottle | PickCharger | 5.9–18.8 | 0.54 | 3.1 |
| PickBolt | PlaceBread | |||||
| Layer | Scope | Contents |
|---|---|---|
| Environment | benchmark-wide | Physics rate 400 Hz and control rate 20 Hz. Solver iterations 24/8, contact offset 1 mm, deterministic solver. Work table geometry and friction. |
| Hand | benchmark-wide | Joint stiffness, armature and joint friction from the vendor model, unscaled; damping from the model’s damping ratio; torque limits from the URDF. Finger-pad contact model ( , 5 mm torsional patch radius). |
| Arm | benchmark-wide | Absolute position of the platform and of the arm roots relative to the work table. Arm start configuration. |
| Task | per task | Objects, their contact materials, masses and inertias. Layout and its randomization ranges. Success conditions and hold durations. Per-episode initialization. |
| Environment layer | |
|---|---|
| Rates | Physics 400 Hz, control 20 Hz (20 physics sub-steps per control step) |
| Solver | TGS, 24 position and 8 velocity iterations, PCM, friction every iteration |
| Contact | Contact offset 1 mm, rest offset 0, bounce threshold 2 m/s |
| Integration | Enhanced determinism on, sleeping disabled, CCD off, gravity 9.81 m/s 2 |
| Materials | Fallback ; table and ground ; restitution 0 throughout |
| Hand layer (Sharpa Hand, 22 joints) | |
| Column | Content |
|---|---|
| observation.state | Robot joint positions and velocities followed by task-specific quantities (object poses, stage flags); dimension per stage in Table 8 |
| action | Absolute joint targets, 7 arm + 22 hand per arm (29 / 58) |
| next.reward | 1 while the success condition holds, else 0 |
| next.success , next.done | Success and episode-end flags |
| observation.environment_state | Flattened poses and joint states of all scene bodies |
| Stages | Obs | Stages | Obs | Stages | Obs |
|---|---|---|---|---|---|
| OpenLunchBox | 85 | PlaceBread | 116 | UnscrewCap | 96 |
| OpenCabinet | 79 | ExtractApple | 85 | TurnNut | 97 |
| OpenDrawer | 79 | AppleToBlender | 88 | FastenScrew | 93 |
| OpenFridge | 74 | LayFlatPanel | 134 | PressButton | 111 |
| CloseFridge | 74 | PickDebris | 88 | ClickMouse | 149 |
| OpenLaptop | 123 | InsertPlug | 96 | TypeKeys | 390 |
| Configuration | PPO | RLPD |
|---|---|---|
| Observations and actions | ||
| Image resolution | , base and wrist camera | , base and wrist camera |
| Proprioception | , , EE pose | , , EE pose, |
| Action | Absolute joint targets, 29-D | Joint-target increments, 29-D |
| Action bound | , demo range | , |
| Policy distribution | Gaussian, | -Gaussian, |
| Configuration | ACT | Diffusion Policy |
|---|---|---|
| Visual observations | Front and wrist RGB images at offsets | Front and wrist RGB images at offsets |
| Image resolution | ||
| Proprioceptive state | Current robot joint positions | Current robot joint positions |
| Action representation | Absolute target joint positions | Absolute target joint positions |
| State/action dimension | 29 (single-arm), 58 (bimanual) | 29 (single-arm), 58 (bimanual) |
| Visual encoder | ImageNet-pretrained ResNet-18 | ImageNet-pretrained ResNet-18 with GroupNorm |
| Stages | Timesteps | Contact ratio | Fingers |
|---|---|---|---|
| Opening and closing | 60–162 | 0.52 | 2.3 |
| OpenLunchBox | 99 (93–109) | 0.78 | 1.6 |
| OpenCabinet | 110 (99–120) | 0.58 | 3.2 |
| OpenDrawer | 103 (91–120) | 0.59 | 3.5 |
| OpenFridge | 162 (134–183) | 0.63 | 3.6 |
| CloseFridge | 60 (56–67) | 0.45 | 1.9 |
| Stages | Path length (m) | Pairwise distance (cm) | Time CV |
|---|---|---|---|
| Opening and closing | |||
| OpenLunchBox | 0.53 | 3.9 | 0.13 |
| OpenCabinet | 0.86 | 4.9 | 0.15 |
| OpenDrawer | 0.60 | 3.9 | 0.20 |
| OpenFridge | 1.11 | 6.9 | 0.21 |
| CloseFridge | 1.38 | 8.7 | 0.13 |
| Stages | Replay success | Contact ratio | Events | Fingers |
|---|---|---|---|---|
| Opening and closing | 98% | 0.52 | 2.1 | 2.3 |
| OpenLunchBox | 99% (110/111) | 0.78 | 2.2 | 1.6 |
| OpenCabinet | 99% (100/101) | 0.58 | 1.9 | 3.2 |
| OpenDrawer | 100% (110/110) | 0.59 | 1.9 | 3.5 |
| OpenFridge | 93% (107/115) | 0.63 | 2.0 | 3.6 |
| CloseFridge | 100% (104/104) | 0.45 | 2.2 | 1.9 |
| Task | Demo-guided RL | Imitation learning | VLA models | |||
|---|---|---|---|---|---|---|
| PPO | RLPD | ACT | DP | GR00T | (LoRA) | |
| Meal preparation | 0 | 0 | 0 | 0 | 0 | 0 |
| Cabinet and pouring | 0 | 0 | 0 | 0 | 0 | 0 |
| Drawer and plug | 0 | 0 | 0 | 0 | 0 | 0 |
| Nut and bolt | 0 | 0 | 0 | 0 | 0 | 0 |
| Office desk | 0 | 0 | 0 | 0 | 0 | 0 |
Long-Horizon Robotic Manipulation
Long-horizon robotic manipulation studies how robots plan, remember, and recover across extended sequences of physical tasks.
See all 459 papers