File size: 8,515 Bytes
f9c7324
8cb339b
 
 
 
f9c7324
8cb339b
f9c7324
8cb339b
 
f9c7324
 
 
 
 
 
8cb339b
 
 
 
 
f9c7324
 
 
 
8cb339b
f9c7324
 
 
 
 
 
8cb339b
 
 
 
 
f9c7324
 
 
8cb339b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f9c7324
 
8cb339b
f9c7324
 
8cb339b
f9c7324
 
 
 
8cb339b
 
 
 
 
 
f9c7324
8cb339b
 
f9c7324
 
 
8cb339b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f9c7324
8cb339b
 
 
 
f9c7324
8cb339b
 
f9c7324
 
8cb339b
f9c7324
 
8cb339b
f9c7324
 
 
8cb339b
f9c7324
8cb339b
f9c7324
8cb339b
 
 
 
 
 
 
 
 
f9c7324
 
8cb339b
 
 
 
 
 
 
 
 
 
 
f9c7324
8cb339b
f9c7324
8cb339b
 
f9c7324
8cb339b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f9c7324
8cb339b
f9c7324
8cb339b
 
f9c7324
8cb339b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
import carb
import math
import torch
import numpy as np
import omni.usd
from omni.kit.scripting import BehaviorScript
from isaacsim.core.prims import Articulation
from omni.isaac.core.prims import XFormPrim
from isaacsim.core.utils.types import ArticulationActions
from pxr import UsdPhysics


def find_prim_path_by_name(prim_name: str, root_path: str = "/") -> str:
    """Recursively find the full path by Prim name"""
    stage = omni.usd.get_context().get_stage()
    prim = stage.GetPrimAtPath(root_path)

    def _find_prim(current):
        if current.GetName() == prim_name:
            return current.GetPath().pathString
        for child in current.GetChildren():
            result = _find_prim(child)
            if result:
                return result
        return None

    found_path = _find_prim(prim)
    if not found_path:
        raise RuntimeError(f"Prim '{prim_name}' not found under {root_path}")
    return found_path


POSSIBLE_ASSET_ROOT_NAMES = ["fridge_E_body_61"]
JOINT_NAMES = ["RevoluteJoint_fridge_left", "RevoluteJoint_fridge_right"]
JOINT_THRESHOLD = 0.4
LIGHT_PRIM_NAME = "fridge_light"


class CupboardControl(BehaviorScript):
    def on_init(self):
        articulation_root_path = None
        for root_name in POSSIBLE_ASSET_ROOT_NAMES:
            try:
                articulation_root_path = find_prim_path_by_name(root_name)
                print(f"Found articulation root: {articulation_root_path}")
                break
            except RuntimeError:
                continue

        if articulation_root_path is None:
            raise RuntimeError(f"Could not find articulation root with names: {POSSIBLE_ASSET_ROOT_NAMES}")

        self.articulation_root_path = articulation_root_path
        self.art = None
        self.initialized = False
        self.joint_names = JOINT_NAMES
        self.joint_indices = None
        self.use_torch = None
        self.joint_threshold = JOINT_THRESHOLD

        self.light_scopes = [XFormPrim(find_prim_path_by_name(LIGHT_PRIM_NAME))]
        self.light_scopes[0].set_visibility(False)

        stage = omni.usd.get_context().get_stage()
        self.joints_info = {}
        for joint_name in self.joint_names:
            joint_prim_path = find_prim_path_by_name(joint_name)
            joint_prim_obj = stage.GetPrimAtPath(joint_prim_path)
            joint = UsdPhysics.RevoluteJoint(joint_prim_obj)
            self.joints_info[joint_name] = {
                "prim": joint,
                "lower_limit": joint.GetLowerLimitAttr().Get(),
                "upper_limit": joint.GetUpperLimitAttr().Get(),
            }

    def on_play(self):
        self._initialize_articulation()

    def on_stop(self):
        self.light_scopes[0].set_visibility(False)

    def on_update(self, current_time: float, delta_time: float):
        if delta_time <= 0:
            return
        if not self.initialized:
            self._initialize_articulation()
            if not self.initialized:
                return
        self.apply_behavior()

    def apply_behavior(self):
        if not self.initialized or self.art is None:
            return
        self.light_control()
        self.damping_stiffness_change()
        self.target_pose_control()

    def _initialize_articulation(self):
        """Initialize the articulation object and get joint information"""
        if self.initialized:
            return

        try:
            self.art = Articulation(self.articulation_root_path)
            if self.art is None:
                return

            # Get joint indices
            self.joint_indices = []
            for joint_name in self.joint_names:
                idx = self.art.get_dof_index(joint_name)
                if idx is not None:
                    self.joint_indices.append(idx)
                else:
                    print(f"Warning: Could not find joint index for {joint_name}")

            if not self.joint_indices:
                print("Warning: No valid joint indices found")
                return

            # Detect if using torch
            positions = self.art.get_joint_positions()
            self.use_torch = isinstance(positions, torch.Tensor)
            print(f"Detected platform: {'Isaac Lab (torch)' if self.use_torch else 'Isaac Sim (numpy/math)'}")

            self.initialized = True
        except Exception as e:
            print(f"Error initializing articulation: {e}")
            self.initialized = False

    def _get_scalar_value(self, value):
        """Convert a value (potentially a torch.Tensor) to a scalar float"""
        if isinstance(value, torch.Tensor):
            return value.item()
        elif isinstance(value, (list, tuple, np.ndarray)):
            return float(value[0] if len(value) > 0 else 0.0)
        else:
            return float(value)

    def light_control(self):
        positions = self.art.get_joint_positions().squeeze()
        left_joint_state = self._get_scalar_value(positions[self.joint_indices[0]])
        right_joint_state = self._get_scalar_value(positions[self.joint_indices[1]])

        doors_open = [
            left_joint_state > math.radians(30),
            right_joint_state > math.radians(30)
        ]

        current_visibility = self.light_scopes[0].get_visibility()
        if any(doors_open):
            if not current_visibility:
                for light in self.light_scopes:
                    light.set_visibility(True)
        else:
            if current_visibility:
                for light in self.light_scopes:
                    light.set_visibility(False)

    def damping_stiffness_change(self):
        positions = self.art.get_joint_positions().squeeze()
        stage = omni.usd.get_context().get_stage()

        for i, joint_name in enumerate(self.joint_names):
            if i >= len(self.joint_indices):
                continue
            joint_idx = self.joint_indices[i]
            joint_state = self._get_scalar_value(positions[joint_idx])

            joint_drive_path = stage.GetPrimAtPath(find_prim_path_by_name(joint_name))
            joint_drive = UsdPhysics.DriveAPI.Get(joint_drive_path, "angular")
            joint_drive.GetStiffnessAttr().Set(self.calculate_stiffness(joint_state, i))

    def calculate_stiffness(self, joint_state, joint_idx):
        """Calculate stiffness value based on joint angle"""
        max_stiffness = 110
        min_stiffness = 2
        decay_rate = 8

        joint_name = self.joint_names[joint_idx]
        threshold_value = abs(math.radians(self.joint_threshold * (self.joints_info[joint_name]['upper_limit'] + self.joints_info[joint_name]['lower_limit'])))
        if abs(joint_state) >= threshold_value:
            return min_stiffness

        stiffness = max_stiffness * math.exp(-decay_rate * abs(joint_state))
        return max(min_stiffness, min(stiffness, max_stiffness))

    def target_pose_control(self):
        positions = self.art.get_joint_positions().squeeze()
        targets = []
        indices = []

        for i, joint_name in enumerate(self.joint_names):
            if i >= len(self.joint_indices):
                continue
            joint_idx = self.joint_indices[i]
            joint_state = self._get_scalar_value(positions[joint_idx])

            lower_limit = self.joints_info[joint_name]['lower_limit']
            upper_limit = self.joints_info[joint_name]['upper_limit']
            threshold_value = math.radians(self.joint_threshold * (upper_limit + lower_limit))

            if abs(lower_limit) <= math.radians(abs(upper_limit)):
                if joint_state > threshold_value:
                    target = math.radians(upper_limit)
                else:
                    target = math.radians(lower_limit)
            else:
                if joint_state < threshold_value:
                    target = math.radians(lower_limit)
                else:
                    target = math.radians(upper_limit)

            targets.append(target)
            indices.append(joint_idx)

        if not targets:
            return

        if self.use_torch:
            device = positions.device
            targets_tensor = torch.tensor(targets, device=device)
            indices_tensor = torch.tensor(indices, device=device, dtype=torch.long)
            action = ArticulationActions(joint_positions=targets_tensor, joint_indices=indices_tensor)
        else:
            action = ArticulationActions(targets, joint_indices=indices)

        self.art.apply_action(action)