Agility Robotics Hits 100,000 Totes Milestone, Firing a Data-Driven Shot in the Humanoid Wars

In the race to commercialize humanoid robots, the battleground is shifting from viral video stunts to the mundane, grinding reality of throughput. Agility Robotics, the Oregon-based maker of the bipedal robot Digit, announced today that its fleet has successfully moved over 100,000 totes at a GXO Logistics facility in Flowery Branch, Georgia.

The milestone represents a significant data point in an industry often criticized for valuing "lab queens" over reliable workers. By quantifying the volume of work performed, Agility is attempting to distinguish itself as the mature, operational choice in a field crowded with newer, flashier competitors.

The Boring, Necessary Truth of Logistics

The announcement focuses heavily on the "unsexy" reality of Third-Party Logistics (3PL). According to the company, the 100,000-tote metric was achieved during live production cycles where Digit was tasked with moving inventory from autonomous mobile robots (AMRs) to conveyors.

This specific workflow highlights Agility’s argument for the humanoid form factor: versatility. While AMRs are efficient at transport, and industrial arms are superior at fixed manipulation, Digit is designed to bridge the gap—handling the "last meter" of material handling that usually requires human dexterity.

This deployment is part of a broader initiative by GXO, which has been testing a "humanoid workforce" consisting of robots from Agility, Apptronik, and Reflex. However, Agility’s release of specific volume metrics suggests they may be further along in validation than their peers.

A War of Numbers

The timing of this data release is likely no coincidence. It comes just one day after competitor Figure released a post-mortem on its own pilot program at BMW. In that report, Figure revealed its F.02 fleet had manipulated roughly 90,000 sheet metal parts over the course of 1,250 hours.

By claiming "over 100,000 totes," Agility effectively places a chip on the table that matches or slightly exceeds Figure’s volume claims.

The subtext here is a simmering tension between the two companies. Earlier this month, the competition spilled over into a public feud on X, where Figure CEO Brett Adcock predicted Agility would be "bankrupt in <12 months" due to "poor engineering choices." Agility’s response today is not a meme, but a metric—signaling that they intend to compete on uptime and reliability rather than social media engagement.

Reliability as the "Killer App"

Agility’s focus on repetition aligns with the views of industry experts who argue that consistency is the only metric that matters for scaling. Dr. Scott Walter, a veteran robotics analyst, recently noted in an interview with Humanoids Daily that "reliability beats novelty" and that "useful work means the robot must be as reliable as a person."

Walter argues that the industry must target "low-hanging fruit"—boring, repetitive tasks like tote handling—before attempting complex dexterity. Agility appears to be executing exactly that playbook.

"If that new automation can’t maintain performance across thousands of cycles and handle varying conditions over an extended deployment, it’s a liability, not an asset," Agility stated in its release, implicitly contrasting its approach with competitors focused on dynamic, acrobatic movement.

The Global Context

Agility is not operating in a vacuum. While the US market sees a rivalry between Figure and Agility, Chinese competitors are moving aggressively toward mass deployment. UBTECH recently claimed the world's first mass delivery of industrial humanoids, shipping hundreds of units to automotive clients.

As the calendar turns toward 2026, the era of the "demo" is effectively over. Whether it is moving 100,000 totes in Georgia or swapping batteries on an assembly line in Shenzhen, the metric for success has officially changed from "can it do it?" to "how many times can it do it without breaking?"