Uni-PNP: Unified Prehensile and Non-prehensile Policy Learning for Generalizable Object Pose Reaching
Authors: Wenxuan Zhu, Yuze Wang, Jiangran Lyu, Jiayi Chen, Yixin Zheng, Mi Yan, Li Yi, Zhizheng Zhang, He Wang
Organizations: Peking University · Galbot · Institute of Automation, Chinese Academy of Sciences · Beijing Academy of Artificial Intelligence · Tsinghua University
Object manipulation often requires combining non-prehensile actions with grasping. However, existing approaches typically develop these capabilities separately and integrate them by sequencing distinct primitives through prescribed intermediate objectives, limiting the coordination of contact motions and grasping. Their integration also remains sensitive to object and gripper geometry and depends on particular environmental supports, complicating real-world deployment. We present Uni-PNP, a unified policy that jointly learns prehensile and non-prehensile manipulation for object pose reaching across diverse objects and grippers. To achieve this, we first pretrain a geometry encoder to predict contact-induced object motion using a contact-dynamics dataset constructed from simulated interactions between diverse objects and grippers. We then train the policy in a two-stage reinforcement learning curriculum from graspable objects to ungraspable peers, encouraging the discovery of non-prehensile interactions and their integration with grasping. Experiments show 76.39% success on hundreds of seen objects with grippers, and 57.91% on unseen peers, outperforming the strongest pose-reaching baseline by 20.52 and 17.79 percentage points, respectively. We also analyze emergent behaviors, including nonprehensile and prehensile repositioning, grasp acquisition that requires grasping after tipping, and table-assisted reorientation while grasping, and transfer to the real-world scenario. Project page available at lgswdn.github.io/Uni-PNP.
Figures & tables
Fig. 2: Overview of Uni-PNP. We pretrain a geometry encoder to predict contact-induced object motion from object–gripper geometry and gripper actions. The pretrained encoder is then frozen during actor–critic policy learning with a two-stage curriculum.
Fig. 3: Selected contact interactions represented in the pretraining dataset with various generated grippers.
Ungraspable objects
Unseen Objaverse objects
Method
Training grippers
Panda
Training grippers
Panda
GOMP*
–
12.33
–
11.04
PointNet (E2E)
37.67
19.43
37.55
16.91
UniCORN (original)
2.64
2.25
4.58
4.78
UniCORN (retrained)
55.87
42.86
50.08
40.12
Point2Vec
50.64
34.63
46.99
32.05
Table 1: Baseline comparison. Uni-PNP outperforms the pose-reaching baselines across object and gripper settings. GOMP* reports success in grasping, lifting, and holding; all other methods report object pose-reaching success.
Fig. 4: Training success across the two-stage curriculum. The vertical line marks the stage transition at 5,000 iterations. UniCORN and P2V denote UniCORN (retrained) and Point2Vec, respectively.
Pretraining
Architecture
Contact
Dynamics
PointNet
10.89
76.39
UniCORN
55.87
55.11
Table 2: Ablation of architecture and pretraining. PointNet with dynamics pretraining is our encoder; UniCORN with contact pretraining corresponds to UniCORN (retrained).
Fig. 5: Emergent prehensile and non-prehensile manipulation skills. Rows show NP only, P only, NP+P, and P+NP, respectively.
Fig. 6: Skill composition in two representative qualitative rollouts. Colored bars indicate manipulation phases.
Fig. 7: Qualitative real-world object pose-reaching sequence. The goal pose is shown transparently.
Skill
Occurrence (%)
Time share (%)
non-prehensile repositioning
76.71
9.00
prehensile repositioning
99.18
21.82
grasp acquisition
100.00
40.70
table-assisted reorientation
18.90
1.53
Table 3: Manipulation skills in successful episodes. Free motion and transitions account for the remaining time.
Non-prehensile manipulation is often used as a preparatory step for robotic grasping, yet existing approaches typically require a predefined target object pose. In practice, however, objects admit multiple graspable configurations and the desired pose is not known in advance. We reformulate non-prehensile manipulation for grasping as optimizing an object centric graspability objective rather than reaching a specific pose. We construct a graspable set from synthesized grasps and define a graspability field that measures how suitable an object configuration is for successful grasp execution. The scalar measure provides a dense learning signal for reinforcement learning and determines when to terminate manipulation. This yields a closed-loop manipulation-to-grasp pipeline driven by a single policy. Experiments in simulation and on a real robot show that the policy reliably reconfigures objects into graspable states and transitions to grasping without external planners or manually specified stopping conditions. The predicted graspability distance correlates with real world grasp success, which indicates that the learned representation captures grasp feasibility of object configurations.
Licheng Zhong, Gim Hee Lee
Department of Computer Science, National University of Singapore, Singapore
In-hand manipulation allows multi-fingered dexterous hands to reconfigure grasped objects without releasing and regrasping them. This improves manipulation efficiency by reducing repeated grasp acquisition and large arm motions. However, most learning-based methods focus on reorientation, continuous rotation, or translation, whereas many tasks require joint control of object position and orientation. We formulate this capability as in-hand 6D object pose reaching: starting from an existing grasp, coordinated finger motions move the object to a palm-relative target pose. We present POISE (Palm-relative Object reaching In SE(3)), a sim-to-real reinforcement learning framework for this task. POISE combines diverse stable-grasp initialization, goal- and geometry-conditioned control, an adaptive 6D goal curriculum, and a compact reward scheme for pose reaching and grasp preservation. In simulation, diverse initialization raises held-out-grasp success from 40.1% to 51.5% and post-drop recovery from 33.8% to 72.9%; the curriculum raises full-range success from 6.2% to 59.5%. On hardware, the grasp-maintenance reward improves three-target sequence success from 20% to 80%. In real-world experiments, POISE reaches successive 6D targets without manual reset across multiple object geometries and wrist orientations, and recovers from external disturbances. To support further research in dexterous manipulation, we will release our code at https://junxiaolin.github.io/poise-website/.
Junxiao Lin, Tianyue Wu, Jie Yin +3
Sharpa. · School of Computer Science, The University of Sydney, Australia. · School of Computing and Data Science, The University of Hong Kong, HKSAR.
Non-prehensile manipulation offers versatile skills for moving and rearranging heavy or bulky objects, particularly when combined with a mobile manipulation platform. However, both model-based and model-free approaches struggle with the complex hybrid dynamics and the sparsity of the contact in these tasks. To address these challenges, we propose a contact-guided exploration strategy implemented within a Multi-Critic Reinforcement Learning (RL) framework. A dedicated exploration critic is trained with a dense contact-seeking reward that guides the end-effector toward meaningful contact points; its influence is progressively decayed to recover a task-optimal policy. We obtain candidate interaction points from a general-purpose grasping algorithm, enabling the exploration mechanism to generalise across various object geometries. We evaluate the approach on multiple tasks, including box pushing, chair transportation, and a dishwasher opening task. Finally, we validate the chair transportation policy through extensive experiments on a quadrupedal mobile manipulator, demonstrating deployable non-prehensile manipulation in the real world.
Simone Tolomei, Mayank Mittal, Franco Angelini +3
Centro di Ricerca E. Piaggio, Dipartimento di Ingegneria dell’Informazione, Universit`a di Pisa, Italy · ETH Zürich, Switzerland · NVIDIA