New Metric Counts Humanoid Robots' Working Hours

The proposed Worken measure could shift robot buying from hardware specs toward measured output, though it is not yet an industry standard.

Updated on Nov. 3, 2026 in Robotics

Isometric editorial illustration of identical battery modules on a steel rack beside a robotic forearm resting on a workbench, in a plain production hall.
AGIBOT introduced Worken on October 29, 2026, a proposed metric counting the effective working hours its humanoid robots deliver, though it is not yet an industry standard. AI Illustration. Upload story photo >

Live Poll

Do you trust large companies' claims about how much their robots actually help?

AGIBOT introduced Worken, a metric for counting the effective intelligent work hours humanoid robots deliver, at its Innovation Day on October 29, 2026. Worken is a single manufacturer's proposal, not an industry standard, so it is best read as an announced measurement approach rather than an established benchmark.

Why it matters

Humanoid robot makers have largely sold on hardware specs, while buyers want to know what the machines produce in real facilities. Worken tries to make that comparison by multiplying active robots, autonomous working hours and an output efficiency ratio, which signals the field's shift toward pricing around measured output.

Worken is computed from active robot count, autonomous working hours and an output efficiency ratio, and counts only robots working in live environments. The cited figures come from company disclosures, including Agility's 98% task accuracy and 65,000 operating hours, and have not been independently audited.

The players

AGIBOT

China-based humanoid robot maker that introduced the Worken productivity metric.

Agility Robotics

Developer of the Digit humanoid, which is deployed in warehouse work and is merging with Churchill Capital Corp XI.

Figure AI

Maker of the Figure 02 humanoid robot, which ran 10-hour shifts at BMW's Spartanburg, S.C. plant for 11 months.

Amazon

Operator of more than 1 million warehouse robots, which claims potential savings of 15 seconds per item through its supply chain systems.

The details

Worken divides a robot's actual output by a task-specific benchmark, such as the acceptable output a human worker would produce, and then aggregates that ratio across the fleet's active hours. The approach converts robot deployments into a single productivity unit that buyers could compare across vendors. Because the formula excludes cost, it measures what robots accomplish rather than what they cost to operate.

Timeline

  1. October 29, 2026: AGIBOT introduced Worken at its Innovation Day.

  2. September 2026: Agility unveiled its Digit 5 humanoid.

  3. November 2, 2026: Churchill Capital Corp XI shareholders approved the merger with Agility Robotics.

  4. November 5, 2026: The combined Agility company is expected to begin trading on Nasdaq.

The Tech Race

Worken enters a field where Agility's Digit at GXO's Georgia warehouse and Figure AI's Figure 02 at BMW's Spartanburg plant are already being measured against operating hours. Each company is supplying its own performance figures, so no shared standard yet exists to compare them.

Buyers of humanoid robots may begin asking vendors for productivity-based pricing rather than equipment purchases, though the source does not say whether Worken will be adopted by others. Availability of any Worken-based pricing is unannounced.

The takeaway

Watch whether other robot makers adopt Worken or publish a competing metric, since a shared measure would let buyers compare humanoid deployments directly. The combined Agility company's Nasdaq debut on November 5, 2026 will also show how investors price measured robot output.

What happens next

The combined Agility Robotics company is expected to begin trading on Nasdaq on November 5, 2026.

Further reading

For broader context on how humanoid machines are moving from demonstrations into factories and warehouses, see the Robotics section.

Source note: This article includes information reported by PYMNTS.

Live Poll

Do you trust large companies' claims about how much their robots actually help?