AGC turns shop-floor speech into structured knowledge with SmartLogger Voice
- AGC and CEC jointly developed SmartLogger® Voice, a speech-recognition data collection system.
- Generative AI converts recorded shop-floor speech into structured data and knowledge-transfer materials.
- Work analysis time was reduced to one-tenth or less compared with stopwatch/video methods.
- AGC plans to expand the system into safety, equipment maintenance and knowledge-management use cases.
What the system does
AGC and Computer Engineering & Consulting (CEC) developed SmartLogger® Voice to capture spoken information on the shop floor, transcribe it, and apply generative AI to organize and analyze the resulting data so tacit know-how and decision criteria become reusable digital assets.
Verification and impact
Site verification at AGC manufacturing locations showed large efficiency gains: work-analysis and knowledge-transfer preparation time fell by 90% or more. The system enabled accumulation of experience-based decision rules and accelerated data-driven decision-making.
Primary use cases
Use cases include streamlining work analysis—replacing stopwatch and video approaches with voice-recorded tasks to speed PDCA cycles—and converting skilled workers' verbal explanations into decision flows, Q&A, and other formats for knowledge transfer.
Technology, features and rollout
AGC developed the voice-to-text and data-structuring software; CEC provided the speech-recognition technology and collection system. Key features include voice-triggered recording, no need for device operation during work, and focused data capture. AGC plans broader rollout into safety, maintenance and knowledge-management as part of its value-creation DX initiatives.
Source: AGC