AGIBOT proposes ‘Worken’ to measure robots’ productive work

AGIBOT has proposed a way to measure how much work robots actually perform in the field. Robotics & Automation News covered the company’s “Worken” on November 10, 2026. In an announcement on October 29, 2026, AGIBOT presented it as a framework that captures the number of active robots, autonomous working hours and productive output.

AGIBOT’s proposal offers a framework combining operational scale, autonomous working hours and output efficiency ratio in response to the problem that deployment numbers alone cannot evaluate work completed in the field. For companies assessing operational performance, the conditions used to calculate the output efficiency ratio and the treatment of human intervention will determine where the metric can be applied.

Worken’s central idea is to represent work performed by multiplying the number of active robots, autonomous working hours and output efficiency ratio, rather than treating the number of robots manufactured or deployed as the outcome itself. Of the three elements, the number of active robots and hours capture the scale of operations, while the output efficiency ratio is positioned to reflect results. AGIBOT has presented this as a proposed framework, not one already established as a common industry standard. Compared with describing a robot’s capabilities using only operating hours or deployment numbers, this approach seeks to bring operational scale, autonomy and work results together in a single measure. For practical use as a comparison metric, however, it will be important to define what counts as “active,” how human intervention is treated in autonomous working hours, and the work conditions used to calculate the output efficiency ratio. If these definitions differ, comparisons across sites or robot models will be difficult even under the same Worken name. The materials provided do not specify the calculation method or how third-party verification would be conducted. In [our coverage of AGIBOT’s operating deployments at the China International Import Expo, where reported cumulative effective intelligent working hours exceeded 62,000](/en/news/agibot-showcases-operational-cases-at-import-expo-reported-effective-i-bfc739), leaderobot.com reported a cumulative figure of more than 62,000 hours as of October 2026. This is a figure reported separately from the Worken proposal, and it has not been confirmed that the two use the same scope or calculation method. How existing operating-hour totals map to Worken’s components is a point to check when assessing the metric’s practicality. It remains unclear whether the framework will be used only for internal operational management or for evaluations across robot models and work environments. Going forward, it will be necessary to clarify how output efficiency ratios are defined for different types of work, how human intervention and failed tasks are treated, and what measurement data will be disclosed. Without these details, Worken will remain an approach that emphasizes productive work, and its effectiveness as a yardstick for comparing the results of different robots cannot be assessed.

AGIBOT presented the formula “Worken = Number of Active Robots × Autonomous Working Hours × Output Efficiency Ratio.” The company describes Worken as “Effective Intelligent Work Hours” and says the name combines “Work” and “Token.” The proposal shifts the focus from the number of robots manufactured or deployed to useful work performed autonomously and consistently.