---
type: "post"
slug: "figure-index-physical-data-infrastructure"
title: "Figure Index Turns Physical Data Collection into Robotics Infrastructure"
description: "Figure is building a global human-contributed video pipeline as a proprietary training-data layer for its Helix humanoid system."
retrieval_nugget: "Figure Index makes diverse physical-world data collection a dedicated infrastructure layer in the humanoid-agent stack. The scale figures come from Figure, and the company has not yet published enough independent evidence to connect dataset growth with general humanoid reliability."
published_at: "2026-08-28"
updated_at: "2026-08-28"
record_date: "2026-08-28"
date_kind: "published_at"
topics: ["figure","robotics","training-data"]
entities: []
source_urls: ["https://figure.ai/news/introducing-index"]
source_format: "primary-source analysis"
editorial_timing: {"lane":"regular_hourly","scheduled_at":"2026-08-31T13:00:00+03:00","real_news_delta":"The launch matters because general-purpose robots cannot rely on internet text alone; they need broad recordings of human interaction with varied objects and environments."}
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:figure-index-physical-data-infrastructure"
status: "published"
visuals: [{"role":"hero","src":"/images/drip/figure-index-physical-data-infrastructure/figure-index-data-pipeline.webp","alt":"Whiteboard flow from human contributors and video uploads through a quality pipeline into a physical dataset and humanoid training stack.","caption":"New Runtime synthesis of Figure's Index data-collection pipeline. Source: https://www.figure.ai/news/introducing-index"}]
editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:0f496f64e2bd14489961e0dc214dc038a8ef230195f4772b1aeb68536a8e8447","reviewed_at":"2026-08-28T06:08:21.337Z","source_evidence_count":1,"verified_claim_count":2,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":2,"implication_count":1,"watch_condition_count":1,"related_record_count":0}
analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"Figure Index makes diverse physical-world data collection a dedicated infrastructure layer in the humanoid-agent stack.","observed_facts":[{"text":"Figure says Index crossed 264,000 app downloads in more than 100 countries and 44,000 weekly active users during a four-month stealth period.","source_urls":["https://figure.ai/news/introducing-index"]},{"text":"Contributors uploaded more than 16 million videos, and Figure says the pipeline was processing thirty minutes of video every second at launch.","source_urls":["https://figure.ai/news/introducing-index"]}],"mechanism":"A consumer contribution app recruits and pays creators, applies quality controls, and turns diverse first-person physical activity into a proprietary dataset for training and improving Helix.","why_now":"The launch matters because general-purpose robots cannot rely on internet text alone; they need broad recordings of human interaction with varied objects and environments.","implications":["Robotics teams should treat collection incentives, consent, labeling, distribution coverage, quality filters, and model feedback loops as core infrastructure decisions."],"evidence_boundary":"The scale figures come from Figure, and the company has not yet published enough independent evidence to connect dataset growth with general humanoid reliability.","watch_conditions":["Revise the thesis when Figure releases dataset composition, consent and quality methods, controlled generalization results, and evidence that additional data improves deployed robot behavior."],"related_records":[],"new_branch_reason":"No existing New Runtime record treats human-contributed physical video as a standalone hardware-agent infrastructure layer."}
routes: {"html":"https://newruntime.com/posts/figure-index-physical-data-infrastructure/","markdown":"https://newruntime.com/posts/figure-index-physical-data-infrastructure.md","json":"https://newruntime.com/posts/figure-index-physical-data-infrastructure.json"}
---

# Figure Index Turns Physical Data Collection into Robotics Infrastructure

## Retrieval answer

Figure Index makes diverse physical-world data collection a dedicated infrastructure layer in the humanoid-agent stack. The scale figures come from Figure, and the company has not yet published enough independent evidence to connect dataset growth with general humanoid reliability.

Figure Index makes diverse physical-world data collection a dedicated infrastructure layer in the humanoid-agent stack.

The launch matters because general-purpose robots cannot rely on internet text alone; they need broad recordings of human interaction with varied objects and environments.
Figure says Index crossed 264,000 app downloads in more than 100 countries and 44,000 weekly active users during a four-month stealth period. Contributors uploaded more than 16 million videos, and Figure says the pipeline was processing thirty minutes of video every second at launch.

A consumer contribution app recruits and pays creators, applies quality controls, and turns diverse first-person physical activity into a proprietary dataset for training and improving Helix.
Robotics teams should treat collection incentives, consent, labeling, distribution coverage, quality filters, and model feedback loops as core infrastructure decisions.

The scale figures come from Figure, and the company has not yet published enough independent evidence to connect dataset growth with general humanoid reliability.
Revise the thesis when Figure releases dataset composition, consent and quality methods, controlled generalization results, and evidence that additional data improves deployed robot behavior.