On 29 October in Shanghai, the humanoid-robot maker AGIBOT introduced “Worken”, a word made from “work” and “token” . Its co-founder Peng Zhihui framed the question like this: robots are entering factories and warehouses, so “how much productive work are they actually delivering?”
His answer: “We want to move beyond measuring how many robots are built or delivered and begin measuring how effectively they can perform useful work, autonomously and consistently, in the real world.”
Questions this article answers
SAQ: the questions you should be asking (and nobody is)
- What is the one finished outcome my AI tools exist to produce?
- Do I know how many of the actions my tools take end in that outcome?
- Am I reporting activity, like calls, messages and tokens, because it is easier to count?
- Who sets the benchmark for “acceptable work”, and is it the same person selling me the tool?
Stop counting your AI tools. Count what they finish.
The video for this article is in production.
What Worken is
AGIBOT calls it “Effective Intelligent Work Hours” and gives a formula :
Worken = Number of Active Robots × Autonomous Working Hours × Output Efficiency Ratio
AGIBOT
The first factor counts robots actually doing tasks, not units built or sitting in inventory. The second counts hours worked autonomously, not hours switched on. The third compares output with a reference benchmark for a defined task, such as the acceptable work a human completes under set standards .
Why it is still a proposal
AGIBOT itself calls Worken an evolving framework, to be refined through real deployments and operating data . Machine Economy noted that the scores shown on stage, such as 14,400 for security-screening guidance at Guangzhou Metro and 2,956.8 for a Longcheer production line, appeared on a slide and are not printed in the written release . A vendor scoring its own robots is a starting point, not an audit.
The same idea outside robotics
Swap “robots” for “AI tools” and the logic carries over. The number that matters is not how many tools you bought or how many calls they made. It is what they completed to a standard someone accepts.
The business lesson
Sources
Sources are listed in their original language.
- 1AGIBOT, “AGIBOT Introduces Worken to Measure Productive Robot Work at AGIBOT Innovation Day 2026,” press release, Shanghai, 29 October 2026. https://www.agibot.com/article/231/detail/131.html
- 2PR Newswire, same AGIBOT release, 29 October 2026. https://www.prnewswire.com/news-releases/agibot-introduces-worken-to-measure-productive-robot-work-at-agibot-innovation-day-2026-302923486.html
- 3Machine Economy, “AGIBOT wants robots scored in “Worken.” The numbers are on a slide,” October 2026. https://machineeconomy.co.uk/briefings/agibot-worken
- 4Monarchy Media, company figures approved for public use (about 8,000 appointments per month; 59% show rate for general-dentistry appointments), October 2026.