The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
plan_version: string
created: string
sr_specs_path: string
content_root: string
profiles_paths: list<item: string>
child 0, item: string
features_paths: list<item: string>
child 0, item: string
requirements_paths: list<item: string>
child 0, item: string
test_dirs: list<item: string>
child 0, item: string
profiles: list<item: struct<id: int64, name: string, version: string, features: list<item: int64>>>
child 0, item: struct<id: int64, name: string, version: string, features: list<item: int64>>
child 0, id: int64
child 1, name: string
child 2, version: string
child 3, features: list<item: int64>
child 0, item: int64
features: list<item: struct<id: int64, name: string, version: string, tests: list<item: int64>, runtime: strin (... 3 chars omitted)
child 0, item: struct<id: int64, name: string, version: string, tests: list<item: int64>, runtime: string>
child 0, id: int64
child 1, name: string
child 2, version: string
child 3, tests: list<item: int64>
child 0, item: int64
child 4, runtime: string
tests: list<item: struct<id: int64, name: string, version: string, test_file: string, engine: struct<tags: (... 112 chars omitted)
child 0, item: struct<id: int64, name: string, version: string, test_file: string, engine: struct<tags: list<item: (... 100 chars omitted)
child 0, id: int64
child 1, name: string
child 2, version: string
child 3, test_file: string
child 4
...
null>
child 0, item: null
discovery_errors: list<item: null>
child 0, item: null
engine_coverage: struct<required_tags: list<item: string>, configured: list<item: struct<name: string, kind: string, (... 56 chars omitted)
child 0, required_tags: list<item: string>
child 0, item: string
child 1, configured: list<item: struct<name: string, kind: string, tags: list<item: string>>>
child 0, item: struct<name: string, kind: string, tags: list<item: string>>
child 0, name: string
child 1, kind: string
child 2, tags: list<item: string>
child 0, item: string
child 2, uncovered: list<item: null>
child 0, item: null
usd_dependency_sha256: struct<assets\gripper_custom\usd\gripper_custom.usda: string, assets\gripper_custom\usd\runnables\ph (... 377 chars omitted)
child 0, assets\gripper_custom\usd\gripper_custom.usda: string
child 1, assets\gripper_custom\usd\runnables\physics\physx.usda: string
child 2, assets\gripper_custom\usd\payloads\gripper_custom_base.usda: string
child 3, assets\gripper_custom\usd\runnables\physics\newton.usda: string
child 4, assets\gripper_custom\usd\payloads\gripper_custom_meshes.usda: string
child 5, assets\gripper_custom\usd\payloads\gripper_custom_physics.usda: string
child 6, assets\gripper_custom\usd\runnables\physics\mounting_physics.usda: string
test_mass_kg_override: double
hardware_payload_rating_claimed: bool
asset_sha256: string
purpose: string
started_utc: string
to
{'purpose': Value('string'), 'asset_sha256': Value('string'), 'usd_dependency_sha256': {'assets\\gripper_custom\\usd\\gripper_custom.usda': Value('string'), 'assets\\gripper_custom\\usd\\runnables\\physics\\physx.usda': Value('string'), 'assets\\gripper_custom\\usd\\payloads\\gripper_custom_base.usda': Value('string'), 'assets\\gripper_custom\\usd\\runnables\\physics\\newton.usda': Value('string'), 'assets\\gripper_custom\\usd\\payloads\\gripper_custom_meshes.usda': Value('string'), 'assets\\gripper_custom\\usd\\payloads\\gripper_custom_physics.usda': Value('string'), 'assets\\gripper_custom\\usd\\runnables\\physics\\mounting_physics.usda': Value('string')}, 'hardware_payload_rating_claimed': Value('bool'), 'started_utc': Value('string'), 'test_mass_kg_override': Value('float64')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 149, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 129, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 544, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 3052, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2559, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2584, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 558, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 440, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 361, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2381, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2309, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
plan_version: string
created: string
sr_specs_path: string
content_root: string
profiles_paths: list<item: string>
child 0, item: string
features_paths: list<item: string>
child 0, item: string
requirements_paths: list<item: string>
child 0, item: string
test_dirs: list<item: string>
child 0, item: string
profiles: list<item: struct<id: int64, name: string, version: string, features: list<item: int64>>>
child 0, item: struct<id: int64, name: string, version: string, features: list<item: int64>>
child 0, id: int64
child 1, name: string
child 2, version: string
child 3, features: list<item: int64>
child 0, item: int64
features: list<item: struct<id: int64, name: string, version: string, tests: list<item: int64>, runtime: strin (... 3 chars omitted)
child 0, item: struct<id: int64, name: string, version: string, tests: list<item: int64>, runtime: string>
child 0, id: int64
child 1, name: string
child 2, version: string
child 3, tests: list<item: int64>
child 0, item: int64
child 4, runtime: string
tests: list<item: struct<id: int64, name: string, version: string, test_file: string, engine: struct<tags: (... 112 chars omitted)
child 0, item: struct<id: int64, name: string, version: string, test_file: string, engine: struct<tags: list<item: (... 100 chars omitted)
child 0, id: int64
child 1, name: string
child 2, version: string
child 3, test_file: string
child 4
...
null>
child 0, item: null
discovery_errors: list<item: null>
child 0, item: null
engine_coverage: struct<required_tags: list<item: string>, configured: list<item: struct<name: string, kind: string, (... 56 chars omitted)
child 0, required_tags: list<item: string>
child 0, item: string
child 1, configured: list<item: struct<name: string, kind: string, tags: list<item: string>>>
child 0, item: struct<name: string, kind: string, tags: list<item: string>>
child 0, name: string
child 1, kind: string
child 2, tags: list<item: string>
child 0, item: string
child 2, uncovered: list<item: null>
child 0, item: null
usd_dependency_sha256: struct<assets\gripper_custom\usd\gripper_custom.usda: string, assets\gripper_custom\usd\runnables\ph (... 377 chars omitted)
child 0, assets\gripper_custom\usd\gripper_custom.usda: string
child 1, assets\gripper_custom\usd\runnables\physics\physx.usda: string
child 2, assets\gripper_custom\usd\payloads\gripper_custom_base.usda: string
child 3, assets\gripper_custom\usd\runnables\physics\newton.usda: string
child 4, assets\gripper_custom\usd\payloads\gripper_custom_meshes.usda: string
child 5, assets\gripper_custom\usd\payloads\gripper_custom_physics.usda: string
child 6, assets\gripper_custom\usd\runnables\physics\mounting_physics.usda: string
test_mass_kg_override: double
hardware_payload_rating_claimed: bool
asset_sha256: string
purpose: string
started_utc: string
to
{'purpose': Value('string'), 'asset_sha256': Value('string'), 'usd_dependency_sha256': {'assets\\gripper_custom\\usd\\gripper_custom.usda': Value('string'), 'assets\\gripper_custom\\usd\\runnables\\physics\\physx.usda': Value('string'), 'assets\\gripper_custom\\usd\\payloads\\gripper_custom_base.usda': Value('string'), 'assets\\gripper_custom\\usd\\runnables\\physics\\newton.usda': Value('string'), 'assets\\gripper_custom\\usd\\payloads\\gripper_custom_meshes.usda': Value('string'), 'assets\\gripper_custom\\usd\\payloads\\gripper_custom_physics.usda': Value('string'), 'assets\\gripper_custom\\usd\\runnables\\physics\\mounting_physics.usda': Value('string')}, 'hardware_payload_rating_claimed': Value('bool'), 'started_utc': Value('string'), 'test_mass_kg_override': Value('float64')}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Articulated robot gripper assets
Prepared by Qingyu Zhang (AlfredZhang98). Static conformance and configured simulation results are documented separately. No NVIDIA collection review or hardware certification is claimed.
Custom gripper hardware project / 自定义夹爪制作项目
GripperOne, maintained by AlfredZhang98, is the hardware construction and control project corresponding to gripper_custom/ in this dataset. It builds on RhinoV2.0_Gripper. Its documented hardware uses a DAMIAO DM-J4310-2EC motor, a Seeed Studio XIAO ESP32-C6 controller and PaXini CP-M3025 tactile finger pads. The project describes tactile-feedback force control and USB, Bluetooth and Wi-Fi interfaces using a shared communication protocol.
GripperOne 是自定义夹爪的制作与控制项目。最新版仓库包含 SolidWorks 机械装配模型、加工与装配说明、物料清单、主控电路板设计和生产文件、控制板固件及电脑端工具。固件是运行在控制板上的程序。项目文档介绍利用触觉反馈调整夹持力，并通过 USB、蓝牙或 Wi-Fi 控制夹爪。
| Resource / 资料 | Contents / 内容 |
|---|---|
| Hardware overview / 硬件与加工说明 | Components, manufacturing and assembly / 结构组成、加工与装配 |
| SolidWorks models / 三维模型 | Mechanical assemblies and parts / 机械装配体与零件 |
| Bill of materials / 物料清单 | Mechanical and electronic parts, specifications and sourcing links / 机械件、电子件、规格和采购链接 |
| Controller board and production files / 主控电路板与生产文件 | Altium schematic and layout, fabrication and assembly files / Altium 原理图、版图、制板与贴片文件 |
| Wiring and power / 接线与供电 | Pins, connectors and power arrangements / 引脚、接口与供电说明 |
| Firmware / 控制板固件 | Software running on the control board; build, flash and configuration / 固件编译、烧录与配置 |
| User manual / 使用手册 | Connections, calibration and PC tools / 连接、标定与电脑端操作 |
| Motor control notes / 电机控制记录 and protocol / 通信协议 | Bench records, parameters and commands / 台架记录、参数与控制命令 |
The project's material-cost estimate is approximately CNY 3000 per unit, as stated in its bill of materials. This is the maintainer's estimate; prices were not independently rechecked for this dataset. 项目物料成本估计约为人民币 3000 元，依据为仓库维护者的物料清单，本任务未重新询价。An English project overview and English versions of the documentation are also available.
Checked against the GripperOne source commit on 2026-10-27. The hardware project's tactile-force control and USB, Bluetooth and Wi-Fi interfaces are outside this dataset's simulation validation. 硬件功能与本数据集已完成的仿真验证范围分别说明，见 validated simulation scope / 仿真验证范围.
Simulation assets and validation
| Asset | Engine | Pass (tests) | Fail (tests) | Error (tests) | Skipped (tests) |
|---|---|---|---|---|---|
| gripper_ctag2f90d | PhysX | 17 | 0 | 0 | 3 |
| gripper_ctag2f90d | Newton | 17 | 0 | 0 | 3 |
| gripper_custom | PhysX | 17 | 0 | 0 | 3 |
| gripper_custom | Newton | 17 | 0 | 0 | 3 |
Both grippers retain articulated physics. The custom gripper includes two copies of the supplied follower-rod STL. The full GLB supplies assembly and appearance references; it is not a single replacement rigid body. CTAG2F90-D appearance follows the photograph: charcoal metal, silver existing hardware and dark silicone pads. The GLB has 19 appearance materials but no mass or inertia fields.
| Asset | Rigid bodies (count) | Revolute joints (count) |
|---|---|---|
| Custom gripper | 7 | 6 |
| CTAG2F90-D | 7 | 6 |
The user confirmed that the measured custom total of 0.8056 kg includes both CNC-machined aluminium-alloy rods. Each rod is estimated at 0.024964965 kg using its closed STL volume and an explicit nominal density of 2700 kg/m3. Alloy grade and separate weights are unknown. Existing masses and inertia eigenvalues are scaled by 0.9380214372, retaining local centers of mass and principal-axis directions. These separate-body values are estimates, not independent measurements.
Separate base revolute joints follow the leader without redundant distal loop constraints. The original approximate parallel-jaw mechanism is retained. Maximum transferred distal-bore mismatch is 0.374417 mm over the 0.55 rad travel. This limitation is not represented as exact mechanism calibration.
| Geometry source | Bore or pivot spacing (mm) |
|---|---|
| Supplied rod STL | 60.000 |
| Original URDF main link | 60.505 |
The interface USD references current base, geometry and neutral-physics layers. PhysX and Newton data are isolated in separate runtime payloads. Enable exactly one runtime; both are Disabled by default. Mounting=Free is the default for parent-robot attachment or free-body tests. Mounting=Fixed adds a world-to-base fixed joint for a bench at the asset origin. When moving or rotating a fixed bench, update the world anchor. Fixed-mode Robot Core requirements pass in separate official configuration validation. Independent MuJoCo asset support is not declared. Delivered asset materials use USD Preview Surface and do not include MDL or OpenPBR. The official display-colour test may use its own OpenPBR probe internally.
CTAG Newton uses 6 N m s/rad passive leader damping. This respects its strictest coupled follower speed with the original effort bound: minimum damping is 10 N m / (2 rad/s) = 5 N m s/rad, with numerical margin. Existing leader and follower speed limits are retained. This is a simulation approximation, not a measured motor curve. All CTAG Newton tests are repeated after this isolated change; unchanged PhysX physics is verified by composition signatures.
Existing effort and speed limits are retained. The following custom settings describe the simulation controller and numerical regularization, not hardware calibration. NewtonMimicAPI is isolated to Newton; the supported PhysX mimic API remains in PhysX.
| Custom parameter | PhysX | Newton | Unit |
|---|---|---|---|
| Leader position gain | 30 | 100 | N m/rad |
| Drive damping | 0.3 | 0.3 | N m s/rad |
| Effort limit | 10 | 10 | N m |
| Original joint numerical armature | 0.01 | 0.1 | kg m2 |
| Added rod numerical armature | 0.001 | 0.001 | kg m2 |
| Passive leader damping | source settings | 12.05 | N m s/rad |
| Other three original joints, passive damping | source settings | 1 | N m s/rad |
These Newton contact settings are numerical parameters, not measured silicone or aluminium constitutive properties. Newton uses an elliptic friction cone. The tangential-gain mapping depends on this cone setting. Results do not establish equivalent behavior under different solver settings.
| Newton parameter | Pad | Structure | Unit |
|---|---|---|---|
| Normal stiffness | 250000 | 10000 | N/m |
| Normal damping | 500 | 300 | N s/m |
| Tangential gain | 10000 | 1000 | N s/m |
| Sliding friction | 1 | 1 | dimensionless |
| Adhesion | 0 | 0 | disabled |
| Solver parameter | Value | Unit |
|---|---|---|
| Outer frequency | 240 | Hz |
| Internal substeps | 4 | count |
| Internal time step | 0.0010416667 | s |
The existing isaaclab Conda environment supplies Python 3.12.14 and Isaac Sim 6.1.0.0, internal build 6.1.0-rc.26, on RTX 5090. Project-scoped dependencies are selected only for task Isaac processes. Shared Conda dependency cleanliness is not asserted. NVIDIA benchmark 2026.8.0.post1 uses documented local patches for world-pin pose, invalid-camera cleanup and applying Newton configuration after each actual test-stage load. Test definitions, criteria, thresholds, measurements and verdicts are unchanged. These are official-suite results with disclosed local adapter configuration, not an unmodified vendor installation.
RTX images use official simready-thumbnail 2026.8.0 camera and lighting with Isaac Sim RTX Real-Time 2.0. Sources are 1024 x 1024 px; packaged thumbnails are 256 x 256 px. The default OVRTX backend was unavailable locally. Renderer substitution is disclosed. Test and render inputs record hashes of every delivered USD layer.
The custom gripper has no calibrated hardware payload or gripping-force rating. The user's estimate below 3 kg is not authored as maxPayload. Default object masses remain as reported by the suite; supplementary trials use explicit 0.50 kg objects. CTAG's 0.8 kg payload comes from the existing engineering table, which separately lists 0.8 kg self-weight. The public manufacturer page independently confirms self-weight only. Simulation tests do not certify hardware performance or NVIDIA collection acceptance.
Download simready-submission.zip for the reproducible project layout with assets/, submission/ and tools/. Individual asset folders are also available directly in this dataset.
See evidence/requirements_compliance.md and each asset’s usd/docs/VALIDATION.json. Manufacturer attribution is retained in CTAG SOURCE.txt. The distributor selected CC BY 4.0; independent manufacturer redistribution permission is not established in the supplied materials.
- Downloads last month
- -