HUMXN Research Labs Launched $50 Million Data Buyback

The Minneapolis startup seeks to license trade records and worker footage to train physical AI and robotics systems.

Updated on Oct. 13, 2026 in Robotics

Isometric editorial illustration of stacked industrial storage containers, representing a large-scale initiative to collect and organize trade data for robotics.
Minneapolis-based HUMXN Research Labs has initiated a $50 million data acquisition program to train AI models for complex robotic labor tasks. AI Illustration. Upload story photo >

Live Poll

Would you sell your business's internal operational data to an AI company for financial compensation?

HUMXN Research Labs has launched a $50 million initiative to acquire real-world trade and manufacturing data from across the United States. The company intends to use these records to train AI models for physical labor and robotics applications.

Why it matters

Acquiring high-quality, real-world operational data is a critical bottleneck for training robots capable of complex physical tasks. By incentivizing the trade industry to contribute, the firm aims to build a proprietary dataset that bridges the gap between digital AI and physical execution.

The initiative offers $20 per hour for approved job-site footage recorded via wearable cameras. Businesses participating must possess at least 10 years of operational history to qualify.

The players

HUMXN Research Labs

A Minneapolis-based startup specializing in the acquisition and licensing of physical-world datasets for AI and robotics training.

The details

HUMXN Research Labs collects information including service records, repair histories, manuals, and training videos to populate its training datasets. The firm performs a manual review process on all submissions to ensure accuracy, while de-identifying sensitive personal information. Acquired data is then prepared and licensed to third-party robotics and artificial intelligence firms building systems for manual labor.

Timeline

  1. October 13, 2026: The data acquisition initiative was officially launched.

The Tech Race

This effort follows the growing industry trend of prioritizing large-scale proprietary data curation for physical AI training. It marks a departure from reliance on synthetic simulation data as the primary engine for training autonomous labor systems.

Trade businesses in Minneapolis and across the country can now apply through the company website to submit records or footage for compensation. The program is currently open, with no cost to apply for local manufacturers or service providers looking to license their existing documentation.

The takeaway

The move suggests that human-centric operational expertise is becoming the most valuable commodity in the race to build autonomous physical systems. Interested parties should watch for the company's future licensing announcements to identify which robotics sectors are the primary recipients of this data.

Further reading

Learn more about the latest innovations in Robotics and how data-driven training is shaping the field.

Live Poll

Would you sell your business's internal operational data to an AI company for financial compensation?