Roboticdata

Name: Commercially captured real-world training and validation data for Physical AI: geospatial and street-level imagery, aerial capture, pedestrian and indoor data, and voxel-based digital-twin maps across 26+ countries, delivered as large licensed datasets plus a capture-and-validation service; includes human-behavior datasets and a US crowdsourced capture network feeding robotaxi geofence definition and autonomous-system validation
Creator: Roboticdata

ROBOTICDATAManage supplier listing

Commercially captured real-world training and validation data for Physical AI: geospatial and street-level imagery, aerial capture, pedestrian and indoor data, and voxel-based digital-twin maps across 26+ countries, delivered as large licensed datasets plus a capture-and-validation service; includes human-behavior datasets and a US crowdsourced capture network feeding robotaxi geofence definition and autonomous-system validation

Open buyer room →

Sample

Brickroad customers

Sample this dataset before you buy it.

Your sourcing agent asks Roboticdata and files the sample in your catalog — private to you.

Start 7-day trialNothing is charged until the trial ends.

Coverage

Universe, instruments and categories.

Industry: Application SoftwareIT Services
Instruments: equities
Categories: Geospatial

Dataset card

Type: not stated
Format: Multimodal — georeferenced imagery and video (street-level and aerial), 3D voxel/mesh digital-twin maps, point-cloud style spatial captures, and labeled human-behavior video sequences
Volume: Petabyte-class: street-level and aerial imagery plus voxel digital twins across 26+ countries. A single country of street-level coverage typically runs tens of petabytes of imagery and derived 3D; the voxel digital-twin platform alone implies multi-petabyte global 3D city models, and the human-behavior video library adds a further multi-petabyte video corpus
Users: N/A — a licensed-dataset vendor, not a user-facing platform; customers are enterprise AI teams (named logos on the homepage include NVIDIA, Amazon Web Services and Verizon), so the relevant count is contracted accounts, likely low double digits
History: 8 years
Update frequency: not stated
Growth: Active and expanding — the organization has rebranded from a mapping product (Voxelmaps) to a Physical AI data company, added a US crowdsourced capture network and a human-behavior data line, and is marketing to AI-lab buyers, which is an aggressive portfolio expansion rather than a mature catalog
Launched: ~2018-2020 for the underlying mapping business (the corporate profile dates the organization to ~2020 and the Voxelmaps product predates the rebrand); the Robotic Data Physical-AI data positioning is a 2024-2025 repositioning
Delivery: not stated
Entity mapping: not stated
Sample: not stated
Point-in-time: not stated
Licence: not stated

supplier index read from their site