Human Archive

Founded by Rushil Agarwal in 2026

Multimodal data provider for robotics and world modeling

Cost

Demo

Time to value

Requires Expertise

You can use Human Archive to build the largest dataset of human sensorimotor intelligence by collecting synchronized multimodal data from thousands of contributors. The system captures motion, vision, depth, tactile feedback, and wrist-perspective video from people performing everyday tasks in homes, hotels, restaurants, and workplaces. This data helps train embodied AI systems to understand how humans move and interact with their environment.

What Human Archive does

Capture synchronized multimodal human activity dataGenerate 3D pose estimations from motion captureRecord tactile feedback during object manipulationCreate dense depth maps from stereo camerasTrack hand movements in real-timeSegment objects and scenes in video recordingsLabel human actions and behaviorsCoordinate data collection across multiple locationsSynchronized capture of motion, vision, depth and tactile dataReal-time hand tracking with finger precisionWrist-mounted perspective recordingDense depth mapping from stereo camerasTactile force feedback measurement3D pose estimation and visualizationObject and scene segmentation labelsData collection across multiple environments