1.1.0 (#132)

* Bug fix and visual change to open_door.
Previously the agent could open without needing to turn the handle.

* Visual change to close_door.

* New overhead camera. Moved over-shoulder cameras.

* Hanger tasks visual updates.

* Added UR5. Resolves #100. Resolves #90.

* [ci skip] Update README.

* Can specify what episodes to load when loading demos from disk.

* Bump PyRep version.

* Fix spawning dummys after each task swap. Resolves #113.

* Workflow fix.

* Update unit test assets.

* Workflow fix.

* Workflow fix.

* Fix typo. Resolves #102.

* Update TV tasks.

* Update clock, box, and shoe tasks.

* Update door tasks.

* Update fridge tasks.

* Jenga and books on bookshelf improvements.

* Experimental action mode. Improved task checks.

* Collision check fix when getting demos.

* Bug fixes in ABS_EE_POSE_PLAN_WORLD_FRAME

* Allow noise to be added to stored demos.

* RGB returned as uint8 rather than float. Point cloud obs added. Ability to get camera info.

* Update tests accounting for RGB in range 0-255.

* Added option for collision checking in ABS_EE_POSE_PLAN_WORLD_FRAME.

* Swapped planner to RRTConnect.

* Path action modes are no longer interrupted mid path.

* Phone on base waypoint update.

* Collision checking swapped to use respondables rather than visuals.

* Improve plan action space when already colliding with object.

* Made stack_wine more difficult by requiring bottle to be ungrasped in rack.

* Fixed gripper not releasing object (#128)

* [skip ci] Update matrix reshape following PyRep update. Resolves #124.

* fixed gripper not releasing object

Co-authored-by: stephen <slj12@ic.ac.uk>

* Update complex_task.md

Convex is better for simulation.

* Revert: Plan action mode can break early.

README.md

@@ -249,6 +249,7 @@ Currently supported arms:
 - Mico arm with Mico gripper `(mico)`
 - Jaco arm with 3-finger Jaco gripper `(jaco)`
 - Sawyer arm with Baxter gripper `(sawyer)`
+- UR5 arm with Robotiq 85 gripper `(ur5)`
 
 You can then swap out the arm using `robot_configuration`:
 

requirements.txt

@@ -3,3 +3,4 @@ Pillow
 pyquaternion
 html-testRunner
 setuptools
+natsort

rlbench/__init__.py

@@ -1,12 +1,12 @@
-__version__ = '1.0.10'
+__version__ = '1.1.0'
 
 import numpy as np
 import pyrep
 
 pr_v = np.array(pyrep.__version__.split('.'), dtype=int)
-if pr_v.size < 4 or np.any(pr_v < np.array([4, 1, 0, 1])):
+if pr_v.size < 4 or np.any(pr_v < np.array([4, 1, 0, 2])):
     raise ImportError(
-        'PyRep version must be greater than 4.1.0.1. Please update PyRep.')
+        'PyRep version must be greater than 4.1.0.2. Please update PyRep.')
 
 
 from rlbench.environment import Environment

rlbench/action_modes.py

@@ -31,18 +31,22 @@ class ArmActionMode(Enum):
     # But does path planning between these points
     ABS_EE_POSE_PLAN_WORLD_FRAME = 8
 
+    # Absolute end-effector pose (position (3) and quaternion (4))
+    # But does path planning between these points (with collision checking)
+    ABS_EE_POSE_PLAN_WORLD_FRAME_WITH_COLLISION_CHECK = 9
+
     # Change in end-effector pose (position (3) and quaternion (4))
     # But does path planning between these points
-    DELTA_EE_POSE_PLAN_WORLD_FRAME = 9
+    DELTA_EE_POSE_PLAN_WORLD_FRAME = 10
 
     # Change in end-effector pose (position (3) and quaternion (4))
     # In the end-effector frame
-    EE_POSE_EE_FRAME = 10
+    EE_POSE_EE_FRAME = 11
 
     # Change in end-effector pose (position (3) and quaternion (4))
     # But does path planning between these points.
     # In the end-effector frame
-    EE_POSE_PLAN_EE_FRAME = 11
+    EE_POSE_PLAN_EE_FRAME = 12
 
     # NOTE: There is no ABS/DELTA mode for the EE_FRAME because ABS == DELTA
 

rlbench/backend/const.py

@@ -6,6 +6,9 @@
 RIGHT_SHOULDER_RGB_FOLDER = 'right_shoulder_rgb'
 RIGHT_SHOULDER_DEPTH_FOLDER = 'right_shoulder_depth'
 RIGHT_SHOULDER_MASK_FOLDER = 'right_shoulder_mask'
+OVERHEAD_RGB_FOLDER = 'overhead_rgb'
+OVERHEAD_DEPTH_FOLDER = 'overhead_depth'
+OVERHEAD_MASK_FOLDER = 'overhead_mask'
 WRIST_RGB_FOLDER = 'wrist_rgb'
 WRIST_DEPTH_FOLDER = 'wrist_depth'
 WRIST_MASK_FOLDER = 'wrist_mask'

rlbench/backend/observation.py

@@ -8,15 +8,23 @@ def __init__(self,
                  left_shoulder_rgb: np.ndarray,
                  left_shoulder_depth: np.ndarray,
                  left_shoulder_mask: np.ndarray,
+                 left_shoulder_point_cloud: np.ndarray,
                  right_shoulder_rgb: np.ndarray,
                  right_shoulder_depth: np.ndarray,
                  right_shoulder_mask: np.ndarray,
+                 right_shoulder_point_cloud: np.ndarray,
+                 overhead_rgb: np.ndarray,
+                 overhead_depth: np.ndarray,
+                 overhead_mask: np.ndarray,
+                 overhead_point_cloud: np.ndarray,
                  wrist_rgb: np.ndarray,
                  wrist_depth: np.ndarray,
                  wrist_mask: np.ndarray,
+                 wrist_point_cloud: np.ndarray,
                  front_rgb: np.ndarray,
                  front_depth: np.ndarray,
                  front_mask: np.ndarray,
+                 front_point_cloud: np.ndarray,
                  joint_velocities: np.ndarray,
                  joint_positions: np.ndarray,
                  joint_forces: np.ndarray,
@@ -25,20 +33,28 @@ def __init__(self,
                  gripper_matrix: np.ndarray,
                  gripper_joint_positions: np.ndarray,
                  gripper_touch_forces: np.ndarray,
-                 wrist_camera_matrix: np.ndarray,
-                 task_low_dim_state: np.ndarray):
+                 task_low_dim_state: np.ndarray,
+                 misc: dict):
         self.left_shoulder_rgb = left_shoulder_rgb
         self.left_shoulder_depth = left_shoulder_depth
         self.left_shoulder_mask = left_shoulder_mask
+        self.left_shoulder_point_cloud = left_shoulder_point_cloud
         self.right_shoulder_rgb = right_shoulder_rgb
         self.right_shoulder_depth = right_shoulder_depth
         self.right_shoulder_mask = right_shoulder_mask
+        self.right_shoulder_point_cloud = right_shoulder_point_cloud
+        self.overhead_rgb = overhead_rgb
+        self.overhead_depth = overhead_depth
+        self.overhead_mask = overhead_mask
+        self.overhead_point_cloud = overhead_point_cloud
         self.wrist_rgb = wrist_rgb
         self.wrist_depth = wrist_depth
         self.wrist_mask = wrist_mask
+        self.wrist_point_cloud = wrist_point_cloud
         self.front_rgb = front_rgb
         self.front_depth = front_depth
         self.front_mask = front_mask
+        self.front_point_cloud = front_point_cloud
         self.joint_velocities = joint_velocities
         self.joint_positions = joint_positions
         self.joint_forces = joint_forces
@@ -47,8 +63,8 @@ def __init__(self,
         self.gripper_matrix = gripper_matrix
         self.gripper_joint_positions = gripper_joint_positions
         self.gripper_touch_forces = gripper_touch_forces
-        self.wrist_camera_matrix = wrist_camera_matrix
         self.task_low_dim_state = task_low_dim_state
+        self.misc = misc
 
     def get_low_dim_data(self) -> np.ndarray:
         """Gets a 1D array of all the low-dimensional obseervations.