from typing import Any, Dict, Union import numpy as np import sapien import torch import math import mani_skill.envs.utils.randomization as randomization from mani_skill.agents.robots import SO100, Fetch, Panda from mani_skill.envs.sapien_env import BaseEnv from mani_skill.envs.tasks.tabletop.pick_cube_cfgs import PICK_CUBE_CONFIGS from mani_skill.sensors.camera import CameraConfig from mani_skill.utils import sapien_utils from mani_skill.utils.building import actors from mani_skill.utils.registration import register_env from mani_skill.utils.scene_builder.table import TableSceneBuilder from mani_skill.utils.structs.pose import Pose #Robomme import matplotlib.pyplot as plt import random from mani_skill.utils.geometry.rotation_conversions import ( euler_angles_to_matrix, matrix_to_quaternion, ) # NOTE: keep wildcard import for legacy helpers that the environment relies on. from .utils import * from .utils.subgoal_evaluate_func import * from .utils.object_generation import * from .utils import reset_panda from .utils.difficulty import normalize_robomme_difficulty from ..logging_utils import logger PICK_CUBE_DOC_STRING = """**Task Description:** A simple task where the objective is to grasp a red cube with the {robot_id} robot and move it to a target goal position. This is also the *baseline* task to test whether a robot with manipulation capabilities can be simulated and trained properly. Hence there is extra code for some robots to set them up properly in this environment as well as the table scene builder. **Randomizations:** - the cube's xy position is randomized on top of a table in the region [0.1, 0.1] x [-0.1, -0.1]. It is placed flat on the table - the cube's z-axis rotation is randomized to a random angle - the target goal position (marked by a green sphere) of the cube has its xy position randomized in the region [0.1, 0.1] x [-0.1, -0.1] and z randomized in [0, 0.3] **Success Conditions:** - the cube position is within `goal_thresh` (default 0.025m) euclidean distance of the goal position - the robot is static (q velocity < 0.2) """ @register_env("PatternLock") class PatternLock(BaseEnv): _sample_video_link = "https://github.com/haosulab/ManiSkill/raw/main/figures/environment_demos/PickCube-v1_rt.mp4" SUPPORTED_ROBOTS = [ "panda", "fetch", "xarm6_robotiq", "so100", "widowxai", ] agent: Union[Panda] goal_thresh = 0.025 cube_spawn_half_size = 0.05 cube_spawn_center = (0, 0) config_hard = { "grid":5, "length":[4,8] } config_easy = { "grid":3, "length":[2,4] } config_medium = { "grid":4, "length":[3,5] } # Combine into a dictionary configs = { 'hard': config_hard, 'easy': config_easy, 'medium': config_medium } def __init__(self, *args, robot_uids="panda_stick", robot_init_qpos_noise=0,seed=0,Robomme_video_episode=None,Robomme_video_path=None, **kwargs): self.achieved_list=[] self.match=False self.after_demo=False self.use_demonstrationwrapper=False self.demonstration_record_traj=False self.robot_init_qpos_noise = robot_init_qpos_noise if robot_uids in PICK_CUBE_CONFIGS: cfg = PICK_CUBE_CONFIGS[robot_uids] else: cfg = PICK_CUBE_CONFIGS["panda"] self.cube_half_size = cfg["cube_half_size"] self.goal_thresh = cfg["goal_thresh"] self.cube_spawn_half_size = cfg["cube_spawn_half_size"] self.cube_spawn_center = cfg["cube_spawn_center"] self.max_goal_height = cfg["max_goal_height"] self.sensor_cam_eye_pos = cfg["sensor_cam_eye_pos"] self.sensor_cam_target_pos = cfg["sensor_cam_target_pos"] self.human_cam_eye_pos = cfg["human_cam_eye_pos"] self.human_cam_target_pos = cfg["human_cam_target_pos"] self.robomme_failure_recovery = bool( kwargs.pop("robomme_failure_recovery", False) ) self.robomme_failure_recovery_mode = kwargs.pop( "robomme_failure_recovery_mode", None ) if isinstance(self.robomme_failure_recovery_mode, str): self.robomme_failure_recovery_mode = ( self.robomme_failure_recovery_mode.lower() ) self.seed = seed normalized_robomme_difficulty = normalize_robomme_difficulty( kwargs.pop("difficulty", None) ) if normalized_robomme_difficulty is not None: self.difficulty = normalized_robomme_difficulty else: # Determine difficulty based on seed % 3 seed_mod = seed % 3 if seed_mod == 0: self.difficulty = "easy" elif seed_mod == 1: self.difficulty = "medium" else: # seed_mod == 2 self.difficulty = "hard" #self.difficulty = "hard" # Use seed to determine number of repetitions (1-5) arbitrarily generator = torch.Generator() generator.manual_seed(seed) self.highlight_starts = {} # Use dictionary to store highlight start time for each button super().__init__(*args, robot_uids=robot_uids, **kwargs) @property def _default_sensor_configs(self): pose = sapien_utils.look_at( eye=self.sensor_cam_eye_pos, target=self.sensor_cam_target_pos ) camera_eye=[0.3,0,0.4] camera_target =[0,0,-0.2] pose = sapien_utils.look_at( eye=camera_eye, target=camera_target ) return [CameraConfig("base_camera", pose, 256, 256, np.pi / 2, 0.01, 100)] @property def _default_human_render_camera_configs(self): pose = sapien_utils.look_at( eye=self.human_cam_eye_pos, target=self.human_cam_target_pos ) return CameraConfig("render_camera", pose, 512, 512, 1, 0.01, 100) def _load_agent(self, options: dict): super()._load_agent(options, sapien.Pose(p=[-0.615, 0, 0])) def _load_scene(self, options: dict): generator = torch.Generator() generator.manual_seed(self.seed) self.table_scene = TableSceneBuilder( self, robot_init_qpos_noise=self.robot_init_qpos_noise ) self.table_scene.build() # Generate 3x3 grid of buttons grid_center = [-0.1, 0] # Grid center position grid_spacing = 0.1 # Spacing between buttons self.buttons_grid = [] self.button_joints_grid = [] avoid = [] button_index = 0 num_rows, num_cols = 5, 8 num_rows, num_cols = self.configs[self.difficulty]["grid"],self.configs[self.difficulty]["grid"] row_center = (num_rows - 1) / 2 col_center = (num_cols - 1) / 2 for row in range(num_rows): # 3 rows (x direction) for col in range(num_cols): # 5 columns (y direction) x_pos = grid_center[0] + (row - row_center) * grid_spacing y_pos = grid_center[1] + (col - col_center) * grid_spacing target_name = f"target_{button_index}" # Create rotation quaternion for vertical target angles = torch.deg2rad(torch.tensor([0.0, 90.0, 0.0], dtype=torch.float32)) rotate = matrix_to_quaternion( euler_angles_to_matrix(angles, convention="XYZ") ) # Build purple and white target target = build_gray_white_target( scene=self.scene, radius=0.02, thickness=0.01, name=target_name, body_type="kinematic", add_collision=False, initial_pose=sapien.Pose(p=[x_pos, y_pos, 0.01], q=rotate), ) self.buttons_grid.append(target) # Note: purple_white_target doesn't have joints, so we append None self.button_joints_grid.append(None) logger.debug(f"Generated target {button_index} at position ({x_pos:.3f}, {y_pos:.3f})") button_index += 1 self.targets_grid = self.buttons_grid # Generate task list to move to each button sequentially tasks = [] # start_end_set = [ # [0, 1, 8, 9], # [6, 7, 14, 15], # [24, 25, 32, 33], # [30, 31, 38, 39] # ] # # Randomly select 2 different sets from start_end_set # set_indices = torch.randperm(len(start_end_set), generator=generator)[:2].tolist() # start_set = start_end_set[set_indices[0]] # end_set = start_end_set[set_indices[1]] # # Randomly select one node from each set # start_idx = torch.randint(0, len(start_set), (1,), generator=generator).item() # end_idx = torch.randint(0, len(end_set), (1,), generator=generator).item() # num_targets = len(self.targets_grid) # node_choices = torch.randperm(num_targets, generator=generator)[:2] # start_node, end_node = node_choices.tolist() # path_nodes, _, _, _ = find_path_0_to_8( # start=start_node, # target=end_node, # R=num_rows, # C=num_cols, # diagonals=True, # generator=generator, # ) # self.selected_buttons = [self.buttons_grid[i] for i in path_nodes] num_targets = len(self.targets_grid) max_attempts = 1000 # Safety limit for attempt in range(max_attempts): node_choices = torch.randperm(num_targets, generator=generator)[:2] start_node, end_node = node_choices.tolist() path_nodes, _, _, _ = find_path_0_to_8( start=start_node, target=end_node, R=num_rows, C=num_cols, diagonals=True, generator=generator, ) length_range = self.configs[self.difficulty]["length"] if length_range[0] <= len(path_nodes) <= length_range[1]: break else: # If we couldn't find a path < 5 after max_attempts, use the last one logger.debug(f"Warning: Could not find path after {max_attempts} attempts") self.selected_buttons = [self.buttons_grid[i] for i in path_nodes] current_target=self.selected_buttons[0] tasks.append({ "func": lambda t=current_target: is_obj_swing_onto(self, obj=self.agent.tcp, target=t), "name": "NO RECORD", "subgoal_segment":f"NO RECORD", "demonstration": True, "failure_func": lambda expected=current_target: self._wrong_button_touch(expected_button=expected), "solve": lambda env, planner, t=current_target: solve_swingonto(env, planner, target=t,record_swing_qpos=True), }) for i, current_target in enumerate(self.selected_buttons[1:]): last_target = self.selected_buttons[i] tasks.append({ "func": lambda t=current_target: is_obj_swing_onto(self, obj=self.agent.tcp, target=t), "name": f"move {direction(current_target, last_target)}", "subgoal_segment":f"move {direction(current_target, last_target)}", "choice_label": f"move {direction(current_target, last_target)}", "demonstration": True, "failure_func": lambda expected=current_target, last=last_target: self._wrong_button_touch(expected_button=expected, last_button=last), "solve": lambda env, planner, t=current_target: solve_swingonto(env, planner, target=t), #"segment":current_target, }) tasks.append({ "func": lambda:reset_check(self,gripper="stick"), "name": "NO RECORD", "subgoal_segment":f"NO RECORD", "demonstration": True, "failure_func": None, "solve": lambda env, planner: [solve_strong_reset(env,planner,gripper="stick")], },) self.selected_buttons = [self.buttons_grid[i] for i in path_nodes] current_target=self.selected_buttons[0] tasks.append({ "func": lambda:reset_check(self,gripper="stick",target_qpos=self.swing_qpos), "name": "NO RECORD", "subgoal_segment":f"NO RECORD", "demonstration": True, "failure_func": None, "solve": lambda env, planner, t=current_target: [solve_strong_reset(env, planner,gripper="stick",action=self.swing_qpos)], }) for i, current_target in enumerate(self.selected_buttons[1:]): last_target = self.selected_buttons[i] tasks.append({ "func": lambda t=current_target: is_obj_swing_onto(self, obj=self.agent.tcp, target=t), "name": f"move {direction(current_target, last_target)}", "subgoal_segment":f"move {direction(current_target, last_target)}", "choice_label": f"move {direction(current_target, last_target)}", "demonstration": False, "failure_func": lambda expected=current_target, last=last_target: self._wrong_button_touch(expected_button=expected, last_button=last), "solve": lambda env, planner, t=current_target: solve_swingonto(env, planner, target=t), #"segment":current_target, }) # Store task list for RecordWrapper use self.task_list = tasks def _initialize_episode(self, env_idx: torch.Tensor, options: dict): with torch.device(self.device): b = len(env_idx) self.table_scene.initialize(env_idx) qpos=reset_panda.get_reset_panda_param("qpos",gripper="stick") self.agent.reset(qpos) def _get_obs_extra(self, info: Dict): return dict() def evaluate(self,solve_complete_eval=False): self.successflag=torch.tensor([False]) self.failureflag = torch.tensor([False]) for idx, button in enumerate(self.buttons_grid): if is_obj_swing_onto(self, obj=self.agent.tcp, target=button):# Only execute when gripper is closed # Update start time to refresh highlight effect when repeatedly triggered self.highlight_starts[idx] =int(self.elapsed_steps[0].item()) # Only record when not recording if self.after_demo==True: if not self.achieved_list or self.achieved_list[-1] is not button: self.achieved_list.append(button) # highlight=reach record, not necessarily target def _to_label(item): name = getattr(item, "name", None) return name if name is not None else str(item) # Backtrack from end of completed button sequence to check if it matches current target sequence exactly achieved_labels = [_to_label(item) for item in self.achieved_list] selected_labels = [_to_label(item) for item in getattr(self, "selected_buttons", [])] remaining = [label for label in selected_labels if label not in achieved_labels] if selected_labels: recent_achieved = achieved_labels[-len(selected_labels):] if len(recent_achieved) == len(selected_labels) and recent_achieved == selected_labels: logger.debug("match success") self.match=True # print(f"achieved_list: {achieved_labels}") # print(f"selected_buttons: {selected_labels}") # print(f"remaining_targets: {len(remaining)}") # Use encapsulated sequence task check function if(self.use_demonstrationwrapper==False):# change subgoal after planner ends during recording if solve_complete_eval==True: allow_subgoal_change_this_timestep=True else: allow_subgoal_change_this_timestep=False else:# during demonstration, video needs to call evaluate(solve_complete_eval), video ends and flag changes in demonstrationwrapper if solve_complete_eval==True or self.demonstration_record_traj==False: allow_subgoal_change_this_timestep=True else: allow_subgoal_change_this_timestep=False all_tasks_completed, current_task_name, task_failed,self.current_task_specialflag = sequential_task_check(self, self.task_list,allow_subgoal_change_this_timestep=allow_subgoal_change_this_timestep) if all_tasks_completed and self.match==False:# Manually set to fail if string match fails logger.debug("match failure") task_failed=True # If task failed, mark as failed immediately if task_failed: self.failureflag = torch.tensor([True]) logger.debug(f"Task failed: {current_task_name}") # If static_check succeeds or all tasks completed, set success flag if all_tasks_completed and not task_failed: self.successflag = torch.tensor([True]) return { "success": self.successflag, "fail": self.failureflag, } def compute_dense_reward(self, obs: Any, action: torch.Tensor, info: Dict): # tcp_to_obj_dist = torch.linalg.norm( # self.agent.tcp_pose.p - self.agent.tcp_pose.p, axis=1 # ) # reaching_reward = 1 - torch.tanh(5 * tcp_to_obj_dist) # reward = reaching_reward*0 reward=torch.tensor([0]) return reward def compute_normalized_dense_reward( self, obs: Any, action: torch.Tensor, info: Dict ): return self.compute_dense_reward(obs=obs, action=action, info=info) / 5 def _wrong_button_touch(self, expected_button, last_button=None): # If button touched by is_obj_swing_onto is neither current expected target nor previous button (debounce), check as error for button in self.buttons_grid: if button is expected_button: continue if last_button is not None and button is last_button: continue if is_obj_swing_onto(self, obj=self.agent.tcp, target=button): return True return False #Robomme def step(self, action: Union[None, np.ndarray, torch.Tensor, Dict]): obs, reward, terminated, truncated, info = super().step(action) # def _to_label(item): # name = getattr(item, "name", None) # return name if name is not None else str(item) # # Backtrack from end of completed button sequence to check if it matches current target sequence exactly # achieved_labels = [_to_label(item) for item in self.achieved_list] # selected_labels = [_to_label(item) for item in getattr(self, "selected_buttons", [])] # remaining = [label for label in selected_labels if label not in achieved_labels] # if selected_labels: # recent_achieved = achieved_labels[-len(selected_labels):] # if len(recent_achieved) == len(selected_labels) and recent_achieved == selected_labels: # print("match success") # self.match=True # print(f"achieved_list: {achieved_labels}") # print(f"selected_buttons: {selected_labels}") # print(f"remaining_targets: {len(remaining)}") # Check if each button is swum onto, and record highlight start time cur_step = int(self.elapsed_steps[0].item()) highlight_position( self, self.agent.tcp.pose.p, start_step=cur_step, end_step=cur_step + 40, cur_step=cur_step, disk_radius=0.005, ) # for idx, button in enumerate(self.buttons_grid): # if is_obj_swing_onto(self, obj=self.agent.tcp, target=button): # # Update start time to refresh highlight effect when repeatedly triggered # self.highlight_starts[idx] = cur_step # # Only record when not recording # if self.after_demo==True: # if not self.achieved_list or self.achieved_list[-1] is not button: # self.achieved_list.append(button) # highlight=reach record, not necessarily target # Apply highlight effect to each triggered button for idx, button in enumerate(self.buttons_grid): start_step = self.highlight_starts.get(idx) if start_step is not None: highlight_obj( self, button, start_step=start_step, end_step=start_step + 40, cur_step=cur_step, disk_radius=0.02*1.002, disk_half_length=0.01*2*1.002, highlight_color=[1.0, 0.0, 0.0, 1.0], use_target_style=True, ) return obs, reward, terminated, truncated, info