Geppetto tracks 1035 robot training datasets including Open X-Embodiment, DROID, and AgiBot World, each with structured metadata on trajectory count, hours of demonstration, robot platforms, license, commercial use, access model, and file format. Every dataset connects to the robots it trains and the professions it affects.
| Dataset | Year | Category | Trajectories | Hours | Platforms | License | Commercial | Access | Format |
|---|---|---|---|---|---|---|---|---|---|
| RoboLoc | 2025 | research-lab | — | — | robot lidar in indoor-outdoor environments | not stated | No | open | — |
| Robot Localization Failure Prediction Dataset | 2025 | research-lab | — | — | simulated LiDAR, odometry and localisation estimates | not stated | No | open | — |
| SEW Multimodal AMR Dataset 2025 | 2025 | research-lab | — | — | autonomous mobile robot, industrial indoor and urban outdoor | not stated | No | open | — |
| TPT-Bench | 2025 | research-lab | — | — | robot with 3D LiDAR, panoramic and RGB-D cameras | not stated | No | open | — |
| Humanoid Atlas — Global OEM & Supplier Supply Chain Database | 2024 | supply-chain-intelligence | — | — | — | MIT | Yes | open | json, csv |
| Fourier ActionNet — Humanoid Manipulation Dataset | 2024 | real-robot | — | — | Fourier GR-2, Fourier GR-1 | Fourier Research License — non-commercial research use | No | gated | hdf5, lerobot |
| RoboCasa — Large-Scale Simulation for Everyday Household Tasks | 2024 | synthetic | — | 2,227 | Franka Panda (simulated) | MIT | Yes | open | json, mp4 |
| Claru — Training Data for Physical AI | 2024 | research-lab | — | — | robot arms, egocentric cameras, autonomous vehicles | Commercial — custom licensing per engagement | Yes | proprietary | mp4, json, parquet |
| RoboMIND — Multi-Embodiment Intelligence Normative Data | 2024 | real-robot | — | — | Franka Emika Panda, Tien Kung humanoid, AgileX Cobot Magic V2.0 | Apache 2.0 | Yes | open | hdf5, mp4, json |
| Humanoid Robot Soccer fall prediction | 2024 | research-lab | — | 37.9 | — | apache-2.0 | Yes | open | — |
| UAV-VisLoc | 2024 | research-lab | — | — | drones plus satellite maps | not stated | No | open | — |
| UEVAVD | 2024 | research-lab | — | — | UAV-view active detection | not stated | No | open | — |
| MMAUD | 2024 | research-lab | — | — | multi-modal anti-UAV sensing | not stated | No | open | — |
| AD4RL | 2024 | research-lab | — | — | real human driving data plus simulator | not stated | No | open | — |
| Beacon | 2024 | research-lab | — | — | intersection video during blackouts | not stated | No | open | — |
| CORP | 2024 | research-lab | — | — | 18 cameras and 9 lidars on campus | not stated | No | open | — |
| HoloVIC | 2024 | research-lab | — | — | holographic intersections plus vehicle | not stated | No | gated | — |
| Off-road racing datasets (Beyond the Mud) | 2024 | research-lab | — | — | trackside photos of off-road motorcycle races | not stated | No | open | — |
| RCooper | 2024 | research-lab | — | — | roadside cameras and lidars | not stated | No | gated | — |
| RoScenes | 2024 | research-lab | — | — | roadside multi-view cameras on highways | not stated | No | gated | — |
| TUMTraf Event | 2024 | research-lab | — | — | roadside event and RGB cameras | not stated | No | gated | — |
| V2AIX | 2024 | research-lab | — | — | ETSI ITS V2X messages from vehicles and RSUs | not stated | No | open | — |
| BadODD | 2024 | research-lab | — | — | driving-view cameras, Bangladesh | not stated | No | open | — |
| Bosch Street Dataset | 2024 | research-lab | — | — | car with imaging radar, lidar, cameras, 360 coverage | not stated | No | open | — |
| CoSEC | 2024 | research-lab | — | — | coaxial stereo event cameras | not stated | No | open | — |
| DIVA (SimGen) | 2024 | research-lab | — | — | web driving videos plus MetaDrive simulation | CC BY-NC-SA 4.0 for annotations | No | open | — |
| DrivingWorld private driving data | 2024 | research-lab | — | — | not stated | not stated | No | open | — |
| EvTTC | 2024 | research-lab | — | — | event camera | not stated | No | open | — |
| HeLiMOS | 2024 | research-lab | — | — | heterogeneous LiDARs | not stated | No | open | — |
| IDD-X | 2024 | research-lab | — | — | multi-view dashcams, India | not stated | No | open | — |
| InterHub | 2024 | research-lab | — | — | interaction events mined from public driving logs | not stated | No | open | — |
| MAN TruckScenes | 2024 | research-lab | — | — | MAN truck, 4 cameras, 6 lidars, 6 radars | not stated | No | open | — |
| MIPD | 2024 | research-lab | — | — | camera, lidar, 4D radar, sound, vibration | not stated | No | open | — |
| MSU-4S | 2024 | research-lab | — | — | MSU vehicle with FLIR cameras, lidar, radar | not stated | No | open | — |
| MUSES | 2024 | research-lab | — | — | camera, lidar, radar, event camera, GNSS | not stated | No | gated | — |
| NAVSIM / NAVSIM v2 | 2024 | research-lab | — | — | nuPlan and OpenScene sensor logs | not stated | No | open | — |
| NSAVP (Novel Sensors for Autonomous Vehicle Perception) | 2024 | research-lab | — | — | novel sensor suite on vehicle | not stated | No | open | — |
| OmniHD-Scenes | 2024 | research-lab | — | — | 6 cameras, 6 4D imaging radars, 128-beam lidar | not stated | No | open | — |
| Open MARS Dataset | 2024 | research-lab | — | — | May Mobility fleet, lidar and surround cameras | not stated | No | open | — |
| ParkingE2E | 2024 | research-lab | — | — | car with surround cameras in garages | not stated | No | open | — |
| ROAD-Waymo | 2024 | research-lab | — | — | Waymo Open videos | not stated | No | gated | — |
| ROADWork | 2024 | research-lab | — | — | vehicle and phone cameras in work zones | not stated | No | open | — |
| RSUD20K | 2024 | research-lab | — | — | driving-view cameras, Bangladesh | not stated | No | open | — |
| SemanticSpray++ | 2024 | research-lab | — | — | camera, lidar, radar on test track | not stated | No | open | — |
| SID (Stereo Image Dataset) | 2024 | research-lab | — | — | ZED stereo camera on vehicle | not stated | No | open | — |
| TUMTraf V2X | 2024 | research-lab | — | — | roadside and vehicle cameras and lidars | not stated | No | gated | — |
| Ultra-AV | 2024 | research-lab | — | — | longitudinal trajectories of automated vehicles | not stated | No | open | — |
| V2X-Real | 2024 | research-lab | — | — | 2 CAVs and 2 smart infrastructure units | not stated | No | open | — |
| V2XPnP Sequential Dataset | 2024 | research-lab | — | — | multiple vehicles and infrastructure | not stated | No | open | — |
| ARCap | 2024 | human-video | — | — | AR headset with mocap gloves | not stated | No | open | — |
| Conversation Partners egocentric dataset | 2024 | human-video | — | — | not stated | not stated | No | open | — |
| EFM3D | 2024 | human-video | — | — | Project Aria glasses | not stated | No | open | — |
| Ego4D-HCap | 2024 | human-video | — | — | Ego4D videos | not stated | No | open | — |
| Ego4D-Sounds | 2024 | human-video | — | — | Ego4D videos | not stated | No | open | — |
| EgoCVR | 2024 | human-video | — | — | large-scale egocentric video sets | not stated | No | open | — |
| EgoMimic | 2024 | human-video | — | — | Project Aria glasses (human) plus low-cost bimanual manipulator (robot) | not stated | No | open | — |
| EgoPoints | 2024 | human-video | — | — | egocentric sequences, EPIC Fields | not stated | No | open | — |
| EgoVid-5M | 2024 | human-video | — | — | egocentric video clips, 1080p | not stated | No | open | — |
| EventEgo3D | 2024 | human-video | — | — | head-mounted fisheye event camera | not stated | No | open | — |
| HEADS-UP | 2024 | human-video | — | — | head-mounted cameras | not stated | No | open | — |
| HO-Cap | 2024 | human-video | — | — | HoloLens headset plus multiple RGB-D cameras | not stated | No | open | — |
| HOI-QA (HOI-Ref) | 2024 | human-video | — | — | Egocentric images | not stated | No | open | — |
| IKEA Ego 3D | 2024 | human-video | — | — | head-mounted device giving ego-view 3D point clouds | not stated | No | open | — |
| MH-VidQA (Multi-Hop EgoQA) | 2024 | human-video | — | — | Long-form egocentric video | not stated | No | open | — |
| MM-Ego QA data / EgoMemoria | 2024 | human-video | — | — | Ego4D videos | not stated | No | open | — |
| PARSE-Ego4D | 2024 | human-video | — | — | Ego4D videos | not stated | No | open | — |
| Smart-Glasses Engagement dataset | 2024 | human-video | — | — | smart glasses | not stated | No | open | — |
| VidEgoThink | 2024 | human-video | — | — | Ego4D videos | not stated | No | open | — |
| Data Scaling Laws in Imitation Learning dataset | 2024 | human-video | — | — | UMI handheld gripper | not stated | No | open | — |
| UMI (Universal Manipulation Interface) | 2024 | human-video | — | — | handheld gripper with wrist-view camera | not stated | No | open | — |
| EgoGen | 2024 | human-video | — | — | Virtual human with embodied camera | not stated | No | open | — |
| MultiGripperGrasp | 2024 | research-lab | — | — | 11 grippers from two-finger to five-finger, incl. human hand | not stated | No | open | — |
| BODex | 2024 | research-lab | — | — | dexterous hands, bilevel optimization synthesis | not stated | No | open | — |
| DexGYSNet (Grasp as You Say) | 2024 | research-lab | — | — | dexterous hand with language guidance | not stated | No | open | — |
| Get a Grip | 2024 | research-lab | — | — | multi-finger hand with RGB, point clouds and NeRFs | not stated | No | open | — |
| Grasp'Em (MultiGrasp) | 2024 | research-lab | — | — | Shadow Hand multi-object grasps | not stated | No | open | — |
| CapGrasp (SemGrasp) | 2024 | research-lab | — | — | human hand grasps with language | not stated | No | open | — |
| Reasoning grasping benchmark | 2024 | research-lab | — | — | GraspNet-1Billion scenes with implicit instructions | not stated | No | open | — |
| DexMimicGen | 2024 | real-robot | — | — | franka_panda, humanoid_other, parallel_jaw | CC-BY-NC-4.0 | No | open | — |
| SubT-MRS | 2024 | human-video | — | 2,000 | drone, multi, spot | CC-BY-4.0 | Yes | open | — |
| HumanPlus | 2024 | human-video | — | 40 | human, unitree_humanoid | unknown | No | open | — |
| DROID | 2024 | real-robot | — | 350 | franka_panda | CC-BY-4.0 | Yes | open | — |
| TartanAviation | 2024 | real-robot | — | 3,374 | aircraft | CC-BY-4.0 | Yes | open | — |
| TartanAir V2 | 2024 | synthetic | — | — | — | CC-BY-4.0 | Yes | open | — |
| Synthetic UAV Flight Trajectories | 2024 | synthetic | — | 21.28 | drone | Apache-2.0 | Yes | open | — |
| TartanDrive 2.0 | 2024 | human-video | — | 7 | offroad_vehicle | MIT | Yes | open | — |
| FMB (Functional Manipulation Benchmark) | 2024 | real-robot | — | 31.3 | franka_panda | CC-BY-4.0 | Yes | open | — |
| Mobile ALOHA | 2024 | real-robot | — | — | aloha | CC-BY-4.0 | Yes | open | — |
| MiraData | 2024 | research-lab | — | — | long-duration video with structured captions | GPL-3.0 | Yes | open | — |
| RONAR (RoboNar) | 2024 | research-lab | — | — | Stretch SE3 mobile manipulator in real homes | cc-by-4.0 | Yes | open | — |
| lerobot/koch_pick_place_* (4 repos) | 2024 | real-robot | — | — | Koch v1.1 arm (robot_type unknown in metadata) | Apache-2.0 | Yes | open | — |
| ManiSkill-HAB demonstrations | 2024 | synthetic | — | — | Fetch mobile manipulator in ManiSkill home rearrangement | mit | Yes | open | — |
| lerobot/unitreeh1_* (4 repos) | 2024 | real-robot | — | — | Unitree H1 (robot_type unknown in metadata) | Apache-2.0 | Yes | open | — |
| Touch-Slide | 2024 | research-lab | — | — | DIGIT sensor slid by hand over toy kitchen objects | CC BY-NC 4.0 | No | open | — |
| Open Spatial Dataset (SpatialRGPT) | 2024 | research-lab | — | — | images with region level 3D scene graph QA | not stated | No | open | — |
| EmbSpatial-Bench | 2024 | research-lab | — | — | embodied 3D scene images, egocentric perspective | cc-by-4.0 | Yes | open | — |
| DHPR | 2024 | research-lab | — | — | dashcam images | not stated | No | open | — |
| Caltech Aerial RGB-Thermal Dataset | 2024 | research-lab | — | — | UAV RGB and thermal cameras | not stated | No | open | — |
| FIReStereo | 2024 | research-lab | — | — | UAS thermal stereo | not stated | No | open | — |
| Maritime vessel tank inspection aerial robot dataset | 2024 | research-lab | — | — | aerial robots inside ship tanks | not stated | No | open | — |