Deployment Evidence

Digit / GXO / SPANX Tote Handling

A public humanoid robot deployment signal involving Agility Robotics Digit in a structured warehouse tote-handling workflow.

This is one of the clearest public signals connecting a humanoid robot to a real warehouse material-handling workflow. It supports serious pilot interest in tote movement, but it does not yet prove procurement-level readiness because key operating data remains undisclosed.

What happened?

GXO and Agility Robotics publicly announced a multi-year Robots-as-a-Service agreement involving Digit in GXO logistics operations. The most buyer-relevant publicly described workflow connects Digit to tote movement in a SPANX fulfillment context, where the robot is described as moving totes from mobile robots or cobots to conveyors.

For Humanoids Watch, this matters because the signal is tied to a named logistics operator, a named humanoid robot, a specific warehouse workflow, and a commercial service structure.

Who is involved?

Humanoid robot vendor

Agility Robotics

Developer of Digit and provider of the service-style deployment model.

Humanoid robot

Digit

Robot associated with the tote-handling workflow.

Logistics operator

GXO

Named external operator connected to the public deployment signal.

Fulfillment customer context

SPANX

The publicly described workflow is tied to a SPANX fulfillment operation.

What is the workflow?

The publicly described workflow centers on moving standardized totes between automation handoff points. In practical terms, this means Digit is not being presented as a general warehouse worker. The relevant workflow is narrower: moving containers from mobile robots or cobots and placing them onto conveyors.

That narrowness matters. It makes the task more measurable and more plausible for an early humanoid pilot than broad, variable warehouse work.

  1. Mobile robot / cobot presents tote
  2. Digit picks or receives tote
  3. Digit moves tote through a defined area
  4. Digit places tote on conveyor
  5. Workflow continues through existing warehouse automation

Evidence assessment

Evidence type Commercial deployment signal

The signal is stronger than a demo or vague partnership because it is connected to a named logistics operator and commercial service agreement.

Evidence quality Strong

The evidence is buyer-relevant, but still depends partly on company announcements and lacks full operating details.

Buyer relevance High for structured pilots

The signal supports pilot interest in warehouse tote and container movement workflows.

Main caveat Operating data incomplete

Public sources do not fully disclose uptime, intervention rate, cycle time, robot count, safety stops, or cost per completed move.

What this evidence supports

Supports

  • A humanoid robot has been publicly connected to a real logistics workflow.
  • The workflow is specific: tote movement between automation handoff points.
  • The use case is narrow enough to evaluate in a structured pilot.
  • The signal is stronger than a vendor-only demo.
  • Warehouse tote handling is a credible early humanoid use case to monitor.
  • Agility Robotics is one of the vendors buyers should evaluate for this workflow.

Does not prove

  • It does not prove broad warehouse labor replacement.
  • It does not prove general-purpose humanoid autonomy.
  • It does not prove procurement-level readiness for Digit.
  • It does not disclose full uptime or reliability history.
  • It does not disclose intervention rate per 100 moves.
  • It does not disclose cycle-time distribution.
  • It does not disclose all-in cost per completed move.
  • It does not prove that the workflow scales easily across sites.

Missing operating data

These are the denominators and operating details buyers still need before using this signal as a benchmark.

Data point Why it matters
Robot count Needed to interpret task volume and fleet productivity
Scheduled operating hours Needed as denominator for utilization
Productive operating hours Shows actual usable work time
Uptime Determines operational reliability
Intervention rate Reveals hidden human labor
Local vs remote interventions Clarifies supervision burden
Average and median cycle time Determines throughput
Cycle-time variance Shows process predictability
Failed moves Reveals reliability and exception load
Safety stops / near misses Critical for deployment design
Cost per completed move Determines economic viability

Buyer interpretation

The Digit / GXO / SPANX signal is meaningful because it connects a humanoid robot to a concrete warehouse workflow rather than a stage demo. Buyers should treat it as evidence that tote movement is worth evaluating, not as proof that humanoid robots are ready for broad warehouse deployment.

The right next step for a similar buyer is not procurement. It is a tightly scoped pilot with defined baseline metrics, uptime targets, intervention logging, safety boundaries, and cost-per-task analysis.

Buyer interpretation: credible structured-pilot signal

Questions buyers should ask before using this as a pilot benchmark

Operating performance

  • How many Digit units were involved?
  • Over how many scheduled operating hours was the workflow measured?
  • What uptime was achieved?
  • What was the intervention rate per 100 tote moves?
  • What were the most common failure modes?

Throughput and economics

  • What was the average and median cycle time?
  • How many successful moves were completed per productive hour?
  • What was the cost per completed move?
  • How did performance compare with the previous manual or automated process?

Supervision and autonomy

  • What percentage of moves completed without human assistance?
  • What percentage required local worker intervention?
  • What percentage required remote support?
  • Were any tasks teleoperated or supervised?

Safety and integration

  • What safety boundaries were required?
  • Could the robot operate near workers?
  • What integration was required with mobile robots, conveyors, WMS/WES, or site systems?
  • What happened during blocked paths, dropped totes, or sensor faults?

Related intelligence

Sources and review status

Last reviewed: May 2026

Customer sourceGXO announcement