Physix FrontierPhysix Frontier · News Briefing Card (enbrief · Nov 3, 2026)
AI adoption in factories stalls: high costs, slow payback, multiple hurdles
KEY FACTS
- After a manufacturing company introduced AI, its human resources department shrank from 12 people to 3.
- Some enterprises spent hundreds of thousands of yuan on servers for AI deployment, but abandoned the project due to adaptation and data security issues.
- Laoganma launched an AI visual quality inspection system in 2025, reducing the defect rate in key processes to 0.02%.
- Laoganma denied claims that AI cut 600 pepper-sorting workers to 48.
- Procter & Gamble partnered with Siemens to deploy AI visual quality inspection, reducing product scrap rate by 10% to 20%.
KEY DATA
0.02%Laoganma key process defect rate
10% to 20%P&G product scrap rate reduction
35%Shougang defect rate reduction
32%CIMC manufacturing cycle shortening
PHYSIX OBSERVATION
AI implementation in factories shows clear stratification: tasks with well-defined boundaries such as quality inspection have yielded results, but production control still struggles to replace automation. Enterprises must beware of treating procurement as cost reduction while ignoring data governance and the shortage of hybrid talent, otherwise investments are easily wasted.
Source: enbrief original report ↗