Update grasp pose generator and pick up action by matafela · Pull Request #289 · DexForce/EmbodiChain

matafela · 2026-06-03T08:30:47Z

Description

Upgrade grasp pose generator and pick up action.

TODO:

grasp pose nms
more grasp pose sampler from different approach direction
return top k but not one grasp pose.
pick up action choose ik valid grasp pose

Type of change

Enhancement (non-breaking change which improves an existing functionality)

Checklist

I have run the black . command to format the code base.
I have made corresponding changes to the documentation
I have added tests that prove my fix is effective or that my feature works
Dependencies have been updated, if applicable.

)

Co-authored-by: chenjian <chenjian@dexforce.com>

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Co-authored-by: chenjian <chenjian@dexforce.com>

Co-authored-by: Jietao Chen <chenjietao@dexforce.com> Co-authored-by: Yueci Deng <dengyueci@qq.com>

Co-authored-by: chenjian <chenjian@dexforce.com>

Co-authored-by: yuanhaonan <yuanhaonan@dexforce.top>

…entation (#247) Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Copilot <copilot@github.com>

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Co-authored-by: chenjian <chenjian@dexforce.com> Co-authored-by: yuecideng <dengyueci@qq.com>

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

…ration (#239) Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Chen Jian <mtfl1996@outlook.com> Co-authored-by: chenjian <chenjian@dexforce.com> Co-authored-by: Copilot <copilot@github.com>

Co-authored-by: chenjian <chenjian@dexforce.com>

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Co-authored-by: Yueci Deng <dengyueci@qq.com>

Co-authored-by: WaferLi <63717327+WaferLi@users.noreply.github.com> Co-authored-by: liwenfeng <liwenfeng@dexforce.top> Co-authored-by: chenjian <chenjian@dexforce.com> Co-authored-by: daojun <lookangela@qq.com> Co-authored-by: Chen Jian <mtfl1996@outlook.com>

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Copilot

Pull request overview

Copilot reviewed 5 out of 5 changed files in this pull request and generated 11 comments.

+            n_poses = grasp_poses.shape[0]
+            init_qpos_repeat = init_qpos[:, None, :].repeat(1, n_poses, 1)
+            grasp_poses_repeat = grasp_poses[None, :, :].repeat(n_envs, 1, 1, 1)
+            ik_success, qpos = self.robot.compute_batch_ik(
+                pose=grasp_poses_repeat,
+                name=self.cfg.control_part,
+                joint_seed=init_qpos_repeat,
+            )
+            if ik_success.sum() == 0:
+                is_success_list.append(False)
+                grasp_xpos_list.append(
+                    torch.zeros(4, 4, dtype=torch.float32, device=self.device)
+                )
+                continue
+            valid_mask = ik_success[0]
+            valid_poses = grasp_poses[valid_mask]
+            valid_costs = grasp_costs[valid_mask]
+            best_idx = torch.argmin(valid_costs)
+            best_grasp_xpos = valid_poses[best_idx]
+            is_success_list.append(True)
+            grasp_xpos_list.append(best_grasp_xpos)


+    def get_valid_grasp_poses(
+        self,
+        obj_poses: torch.Tensor,
+        approach_direction: torch.Tensor = torch.tensor(
+            [0, 0, -1], dtype=torch.float32
+        ),
+    ) -> list[tuple[torch.Tensor, torch.Tensor]]:
+        if self.generator is None:
+            self._init_generator()
+        results = []


+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


@@ -698,6 +718,67 @@ def get_grasp_poses(
        center_cost = center_distance / center_distance.max()
        length_cost = 1 - valid_open_lengths / valid_open_lengths.max()
        total_cost = 0.3 * angle_cost + 0.3 * length_cost + 0.4 * center_cost


+    n_deviated_approach_directions: int = 4
+    """Number of approach directions with evenly deviated angles when sampling grasp poses."""


+def pose_nms_indices(
+    poses: torch.Tensor,
+    angle_th: float = np.pi / 36,
+    dist_th: float = 0.003,
+    preserve_order: bool = False,
+) -> torch.Tensor:
+    """Return pose indices after removing poses that are too close.


        is_success, grasp_pose, open_length = grasp_generator.get_grasp_poses(
            obj_pose,
            approach_direction,
            visualize_collision=False,
-            visualize_pose=False,
+            visualize_pose=True,
        )


+        is_success = torch.tensor(is_success_list, device=self.device)
+        grasp_xpos = torch.stack(grasp_xpos_list, dim=0)
+
+        return is_success, grasp_xpos


Co-authored-by: Cursor <cursoragent@cursor.com>

Copilot

Pull request overview

Copilot reviewed 5 out of 5 changed files in this pull request and generated 10 comments.

Comments suppressed due to low confidence (1)

embodichain/toolkits/graspkit/pg_grasp/antipodal_generator.py:786

get_grasp_poses() is annotated to return best_open_length as a float, but indexing a tensor returns a 0-d torch.Tensor. This can lead to downstream type/device issues (e.g., torch.tensor(open_length_list, device=...) in AntipodalAffordance.get_best_grasp_poses). Return a Python float via .item() (or update the annotation to torch.Tensor).

        best_open_length = valid_open_lengths[best_idx]
        if visualize_pose:

+            n_poses = grasp_poses.shape[0]
+            init_qpos_repeat = init_qpos[:, None, :].repeat(1, n_poses, 1)
+            grasp_poses_repeat = grasp_poses[None, :, :].repeat(n_envs, 1, 1, 1)
+            ik_success, qpos = self.robot.compute_batch_ik(
+                pose=grasp_poses_repeat,
+                name=self.cfg.control_part,
+                joint_seed=init_qpos_repeat,
+            )
+            if ik_success.sum() == 0:
+                is_success_list.append(False)
+                grasp_xpos_list.append(
+                    torch.zeros(4, 4, dtype=torch.float32, device=self.device)
+                )
+                continue
+            valid_mask = ik_success[0]
+            valid_poses = grasp_poses[valid_mask]
+            valid_costs = grasp_costs[valid_mask]
+            best_idx = torch.argmin(valid_costs)
+            best_grasp_xpos = valid_poses[best_idx]
+            is_success_list.append(True)
+            grasp_xpos_list.append(best_grasp_xpos)
+        is_success = torch.tensor(is_success_list, device=self.device)
+        grasp_xpos = torch.stack(grasp_xpos_list, dim=0)
+
+        return is_success, grasp_xpos


            )
        obj_poses = semantics.entity.get_local_pose(to_matrix=True)


+    def get_valid_grasp_poses(
+        self,
+        obj_poses: torch.Tensor,
+        approach_direction: torch.Tensor = torch.tensor(
+            [0, 0, -1], dtype=torch.float32
+        ),
+    ) -> list[tuple[torch.Tensor, torch.Tensor]]:
+        if self.generator is None:
+            self._init_generator()
+        results = []


+    n_deviated_approach_directions: int = 4
+    """Number of approach directions with evenly deviated angles when sampling grasp poses."""


+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


+        TODO:
+            1. Support Top-k grasp poses selection.
+            2. Support more selection criteria.
+


+        # TODO: too slow
+        # # remove near grasp poses using non-maximum suppression
+        # nms_indices = pose_nms_indices(
+        #     valid_grasp_poses,
+        #     angle_th=np.pi / 18,


Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

Copilot

Pull request overview

Copilot reviewed 5 out of 5 changed files in this pull request and generated 11 comments.

+    n_deviated_approach_directions: int = 4
+    """Number of approach directions with evenly deviated angles when sampling grasp poses."""


+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


+        grasp_poses_result = semantics.affordance.get_valid_grasp_poses(
            obj_poses=obj_poses, approach_direction=self.approach_direction
        )


+    def get_valid_grasp_poses(
+        self,
+        obj_poses: torch.Tensor,
+        approach_direction: torch.Tensor = torch.tensor(
+            [0, 0, -1], dtype=torch.float32
+        ),
+    ) -> list[tuple[torch.Tensor, torch.Tensor]]:
+        if self.generator is None:
+            self._init_generator()
+        results = []
+        for i, obj_pose in enumerate(obj_poses):
+            is_success, grasp_poses, _, costs = self.generator.get_valid_grasp_poses(
+                obj_pose, approach_direction
+            )
+            if not is_success:
+                logger.log_warning(
+                    f"Failed to find valid grasp poses for {i}-th object."
+                )
+            results.append((grasp_poses, costs))
+        return results


+        for i in range(n_envs):
+            n_pose = grasp_poses_result[i][0].shape[0]
+            grasp_xpos_padding[i, :n_pose] = grasp_poses_result[i][0]
+            grasp_cost_padding[i, :n_pose] = grasp_poses_result[i][1]
+            # padding with the first grasp pose, which is usually the best one, to ensure that the padded grasp poses are valid for IK computation, although they may not be optimal.
+            grasp_xpos_padding[i, n_pose:] = grasp_poses_result[i][0][0]
+            grasp_cost_padding[i, n_pose:] = grasp_poses_result[i][1][0]
+


+        grasp_cost_masked = torch.where(ik_success, grasp_cost_padding, 10000.0)
+        best_cost, best_idx = grasp_cost_masked.min(dim=1)
+        is_success = best_cost < 9999.0  # usually cost < 1.0
+        best_grasp_xpos = grasp_xpos_padding[
+            torch.arange(n_envs, device=self.device), best_idx
+        ]
+
+        return is_success, best_grasp_xpos


+        n_max_pose = 0
+        for result in grasp_poses_result:
+            n_pose = result[0].shape[0]
+            if n_pose > n_max_pose:
+                n_max_pose = n_pose
+


Copilot

Pull request overview

Copilot reviewed 9 out of 9 changed files in this pull request and generated 12 comments.

            if max_velocity is not None:
                drive_args["max_velocity"] = max_velocity[i].cpu().numpy()
            if friction is not None:
                drive_args["joint_friction"] = friction[i].cpu().numpy()
-            if armature is not None:
-                drive_args["armature"] = armature[i].cpu().numpy()
            self._entities[env_idx].set_drive(**drive_args)


+    def get_valid_grasp_poses(
+        self,
+        obj_poses: torch.Tensor,
+        approach_direction: torch.Tensor = torch.tensor(
+            [0, 0, -1], dtype=torch.float32
+        ),
+    ) -> list[tuple[torch.Tensor, torch.Tensor]]:
+        if self.generator is None:
+            self._init_generator()
+        results = []
+        for i, obj_pose in enumerate(obj_poses):
+            is_success, grasp_poses, _, costs = self.generator.get_valid_grasp_poses(
+                obj_pose, approach_direction
+            )


        if self._hit_point_pairs is None:
            logger.log_warning(
                "No antipodal point pairs available. "
                "Call generate() or annotate() first."
            )
-            return False, torch.eye(4, device=self.device), 0.0
+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


        if valid_mask.sum() == 0:
            logger.log_warning("No valid antipodal pairs after angle filtering.")
-            return False, torch.eye(4, device=self.device), 0.0
+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


        if is_colliding.logical_not().sum() == 0:
-            logger.log_warning("No valid antipodal pairs after angle filtering.")
-            return False, torch.eye(4, device=self.device), 0.0
+            logger.log_warning("No valid antipodal pairs after collision filtering.")
+            return (
+                False,
+                torch.eye(4, device=self.device),
+                0.0,
+                torch.zeros(1, device=self.device),
+            )


+        n_max_pose = 0
+        for result in grasp_poses_result:
+            n_pose = result[0].shape[0]
+            if n_pose > n_max_pose:
+                n_max_pose = n_pose
+
+        grasp_xpos_padding = torch.zeros(
+            (n_envs, n_max_pose, 4, 4), dtype=torch.float32, device=self.device
+        )


+        for i in range(n_envs):
+            n_pose = grasp_poses_result[i][0].shape[0]
+            grasp_xpos_padding[i, :n_pose] = grasp_poses_result[i][0]
+            grasp_cost_padding[i, :n_pose] = grasp_poses_result[i][1]
+            # padding with the first grasp pose, which is usually the best one, to ensure that the padded grasp poses are valid for IK computation, although they may not be optimal.
+            grasp_xpos_padding[i, n_pose:] = grasp_poses_result[i][0][0]
+            grasp_cost_padding[i, n_pose:] = grasp_poses_result[i][1][0]
+


+import argparse
+import numpy as np
+import time
+import torch
+


+    return interp_trajectory
+
+
+def test_grasp_pose_generator():


+        # TODO: too slow
+        # # remove near grasp poses using non-maximum suppression
+        # nms_indices = pose_nms_indices(
+        #     valid_grasp_poses,
+        #     angle_th=np.pi / 18,
+        #     dist_th=0.01,
+        # )
+        # valid_grasp_poses = valid_grasp_poses[nms_indices]
+        # valid_open_lengths = valid_open_lengths[nms_indices]
+        # valid_centers = valid_centers[nms_indices]
+


Copilot

Pull request overview

Copilot reviewed 4 out of 4 changed files in this pull request and generated 8 comments.

+    start = time.perf_counter()
+    is_success, grasp_poses, open_lengths, total_cost = generator.get_valid_grasp_poses(
+        object_pose=object_pose,
+        approach_direction=approach_direction,
+        visualize_collision=False,
+    )


+    return sim
+
+
+def create_robot(sim: SimulationManager, position=[0.0, 0.0, 0.0]) -> Robot:


+def create_mug(sim: SimulationManager):
+    mug_cfg = RigidObjectCfg(
+        uid="table",
+        shape=MeshCfg(
+            fpath=get_data_path("CoffeeCup/cup.ply"),
+        ),


+    else:
+        close = ((reference_rotations @ target_rotations.T - 1.0) * 0.5).clamp_(
+            min=-1.0, max=1.0
+        ) > rotation_cosine_th
+


+    rotation_always_close = angle_th > math.pi
+    rotation_cosine_th = math.cos(float(angle_th)) if not rotation_always_close else 0.0


+def pose_nms_indices(
+    poses: torch.Tensor,
+    angle_th: float = np.pi / 36,
+    dist_th: float = 0.003,
+    preserve_order: bool = False,
+    chunk_size: int = _POSE_NMS_CHUNK_SIZE,
+) -> torch.Tensor:
+    """Return pose indices after removing poses that are too close.


+import argparse
+import numpy as np
+import time
+import torch
+
+from embodichain.lab.sim import SimulationManager, SimulationManagerCfg
+from embodichain.lab.sim.objects import Robot, RigidObject
+from embodichain.lab.sim.utility.action_utils import interpolate_with_distance
+from embodichain.lab.sim.shapes import MeshCfg
+from embodichain.lab.sim.solvers import PytorchSolverCfg
+from embodichain.data import get_data_path
+from embodichain.lab.gym.utils.gym_utils import add_env_launcher_args_to_parser
+from embodichain.utils import logger
+from embodichain.lab.sim.cfg import (
+    RenderCfg,
+    JointDrivePropertiesCfg,
+    RobotCfg,
+    LightCfg,
+    RigidBodyAttributesCfg,
+    RigidObjectCfg,
+    URDFCfg,
+)
+from embodichain.toolkits.graspkit.pg_grasp.antipodal_generator import (
+    GraspGenerator,
+    GraspGeneratorCfg,
+    AntipodalSamplerCfg,
+)
+from embodichain.toolkits.graspkit.pg_grasp.gripper_collision_checker import (
+    GripperCollisionCfg,
+)


+def test_grasp_pose_generator():
+
+    sim = initialize_simulation()
+    robot = create_robot(sim, position=[0.0, 0.0, 0.0])
+    mug = create_mug(sim)


yuecideng and others added 30 commits April 12, 2026 14:58

Docs: rewrite install guide and make lerobot a required dependency (#227

8d20b8d

)

Update cobotmagic arm asset. (#228)

d2a8dad

Co-authored-by: chenjian <chenjian@dexforce.com>

Fix opw solver (#229)

987d04f

Co-authored-by: chenjian <chenjian@dexforce.com>

Fix crashing when no grasp pose found. (#232)

80368bd

Co-authored-by: chenjian <chenjian@dexforce.com>

add rl benchmark (#231)

59021a4

Co-authored-by: chenjian <chenjian@dexforce.com>

Enhance workspace analyzer computational efficiency (#230)

d860435

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Update CobotMagic default safe margin (#235)

d52ea77

Co-authored-by: chenjian <chenjian@dexforce.com>

Fix demo action shape normalization and control-part mapping (#237)

d5bf930

Refine URDF assembly component prefixes and name casing policy (#236)

3bb2592

Co-authored-by: Jietao Chen <chenjietao@dexforce.com> Co-authored-by: Yueci Deng <dengyueci@qq.com>

Refactor: benchmark (#233)

50843a5

Co-authored-by: chenjian <chenjian@dexforce.com>

Update pytorch kinematic solver benchmark param (#240)

e0e16ae

Co-authored-by: chenjian <chenjian@dexforce.com>

fix plan_trajectory (#242)

0497711

Co-authored-by: yuanhaonan <yuanhaonan@dexforce.top>

docs: add NVIDIA driver guide and mesh loading tutorial (#245)

ccce98a

Add multi-version documentation build support (#234)

386dc66

Fix README docs link (#246)

9fb2204

docs: add AI coding agent skills and cross-reference throughout docum…

06258f1

…entation (#247) Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Copilot <copilot@github.com>

fix: correct the link of website (#249)

72e9731

docs: upgrade README badges and add auto-sync to introduction.rst (#250)

c824c86

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

upgrade pytorch kinematics (#244)

8e92c08

Co-authored-by: chenjian <chenjian@dexforce.com> Co-authored-by: yuecideng <dengyueci@qq.com>

chore: upgrade black to 26.3.1 and fix code style (#251)

72b5bb9

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Improve API reference detail and coverage (#252)

fc54598

Fix pytorch solver qpos mapping (#253)

c5e6f00

Co-authored-by: chenjian <chenjian@dexforce.com>

docs: Remove AI coding agent skills references (#254)

23a1208

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

docs: add data generation tutorial for synthesized data pipeline (#238)

076fc65

Co-authored-by: Yueci Deng <dengyueci@qq.com>

Update skills structure and roadmap docs (#257)

f81b8a6

Adapt dexsim v0.4.0 (#226)

9e34ec1

Co-authored-by: WaferLi <63717327+WaferLi@users.noreply.github.com> Co-authored-by: liwenfeng <liwenfeng@dexforce.top> Co-authored-by: chenjian <chenjian@dexforce.com> Co-authored-by: daojun <lookangela@qq.com> Co-authored-by: Chen Jian <mtfl1996@outlook.com>

docs: add uv installation support and update dependencies (#258)

72ffe43

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Fix PyPI release workflow (#259)

c6ce00f

docs: improve navigation, cross-references, and guides (#262)

e11eb4e

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

add top k

31d090a

Copilot AI review requested due to automatic review settings June 4, 2026 07:42

Copilot started reviewing on behalf of matafela June 4, 2026 07:42 View session

Copilot AI reviewed Jun 4, 2026

View reviewed changes

yuecideng requested changes Jun 4, 2026

View reviewed changes

Comment thread embodichain/lab/sim/atomic_actions/actions.py Outdated

Comment thread embodichain/utils/math.py

Comment thread embodichain/toolkits/graspkit/pg_grasp/antipodal_generator.py

yuecideng and others added 4 commits June 5, 2026 00:19

Improve installation guide and deduplicate gensim install docs (#293)

044877e

Co-authored-by: Cursor <cursoragent@cursor.com>

Fix recording for PourWater (#292)

1819048

fix large memory error

b04b343

update

a835741

Copilot AI review requested due to automatic review settings June 9, 2026 02:46

Copilot started reviewing on behalf of matafela June 9, 2026 02:46 View session

Copilot AI reviewed Jun 9, 2026

View reviewed changes

yuecideng and others added 3 commits June 9, 2026 18:22

Auto-select default renderer based on GPU (#294)

cb3240b

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

Merge branch 'main' into cj/upgrade-grasp-generator

dc9acbf

update

96a17d5

Copilot AI review requested due to automatic review settings June 10, 2026 08:56

Copilot started reviewing on behalf of matafela June 10, 2026 08:56 View session

Copilot AI reviewed Jun 10, 2026

View reviewed changes

update

a912e5c

Copilot AI review requested due to automatic review settings June 12, 2026 07:18

Copilot started reviewing on behalf of skywhite1024 June 12, 2026 07:19 View session

Copilot AI reviewed Jun 12, 2026

View reviewed changes

matafela force-pushed the cj/upgrade-grasp-generator branch from f20163f to a912e5c Compare June 12, 2026 07:50

yuecideng mentioned this pull request Jun 12, 2026

Add upright atomic action #302

Open

yuecideng force-pushed the main branch from bbe65f1 to f50efc3 Compare June 15, 2026 04:10

Merge branch 'main' into cj/upgrade-grasp-generator

d412dd7

Copilot AI review requested due to automatic review settings June 15, 2026 04:24

Copilot started reviewing on behalf of yuecideng June 15, 2026 04:25 View session

yuecideng closed this Jun 15, 2026

Copilot AI reviewed Jun 15, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Update grasp pose generator and pick up action#289

Update grasp pose generator and pick up action#289
matafela wants to merge 65 commits into
mainfrom
cj/upgrade-grasp-generator

matafela commented Jun 3, 2026 •

edited by skywhite1024

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants

		n_deviated_approach_directions: int = 4
		"""Number of approach directions with evenly deviated angles when sampling grasp poses."""

		)
		obj_poses = semantics.entity.get_local_pose(to_matrix=True)

		return sim


		def create_robot(sim: SimulationManager, position=[0.0, 0.0, 0.0]) -> Robot:

		rotation_always_close = angle_th > math.pi
		rotation_cosine_th = math.cos(float(angle_th)) if not rotation_always_close else 0.0

Conversation

matafela commented Jun 3, 2026 • edited by skywhite1024 Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Type of change

Checklist

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants

matafela commented Jun 3, 2026 •

edited by skywhite1024

Loading