IT Operations Automation · Real-time Analytics
Pegatron needed to improve factory planning, quality control, and operational efficiency across large manufacturing sites, but manual line design, fragmented agent development, scarce defect data, and labor-intensive compliance checks limited how quickly it could scale industrial AI.
Reported impact
Pegatron reports a 40 percent drop in new-factory construction time, 7 percent lower labor cost per assembly line, 67 percent fewer defects, 400 percent faster AI-agent development, 67 percent shorter visual-inspection model deployment timelines, 10 percent less operational effort, and an estimated 15 percent cut in redundant factory assets through manager-agent orchestration.
Pegatron built PEGAVERSE as a combined digital-twin and AI-factory platform to optimize manufacturing before and during production. The system uses NVIDIA Omniverse, Isaac Sim, and the NVIDIA Metropolis Video Search and Summarization blueprint to model production lines, simulate bottlenecks, and run visual AI agents over live factory data. Pegatron also created an internal AI factory on NVIDIA DGX and NVIDIA AI Enterprise so teams can train and deploy LLM- and VLM-based agents for assembly guidance, warehouse operations, machine-health monitoring, and quality inspection instead of building isolated tools. VSS-powered agents monitor assembly steps in real time and issue corrective alerts, while NVIDIA physical AI skill for Defect Image Generation, NVIDIA Cosmos, and Isaac Sim generate synthetic defect data to improve visual-inspection models. Pegatron is extending the platform with NVIDIA Factory Operations Blueprint and NemoClaw so a factory manager agent can orchestrate specialized agents, robots, and workflows across the plant.
Digital-twin plans and live factory camera or sensor data flow into PEGAVERSE, Omniverse and Isaac Sim simulate or enrich operations, VSS-based agents analyze assembly and defect conditions, and FOX-orchestrated manager agents dispatch insights or actions back to workers, robots, and factory systems.
By augmenting the assembly process with this AI agent, Pegatron is seeing a 7% reduction in labor costs per assembly line and a 67% decrease in defect rates.
Pegatron has accelerated the development of AI agents across the company by 400% over the last four years.
Pegatron reduced AI visual inspection model deployment timelines by 67%, while operational effort decreased by 10%.
This is expected to help reduce construction time for a new factory by 40%.
With the factory manager agent, Pegatron can orchestrate robot utilization more efficiently and reduce the need for expensive standby equipment, with an estimated 15% reduction in asset redundancy costs.
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
IBM Sample May 12, 2026 Suzhou Universal Suzhou Universal needed a faster way to analyze production quality and adjust schedules because MES, ERP, APS, and quality systems were disconnected, cross-system queries were slow, and manual coordination created… 70% reduction in scheduling errors Agentic AI Platform · IBM watsonx Orchestrate 
Microsoft Sample Nov 18, 2024 ABB "For instance, the Genix continuous emission monitoring system can observe and analyze up to 110 different parameters across multiple devices. Based on this data, the system will deliver actionable insights on Genix… "For instance, the Genix continuous emission monitoring system can observe and analyze up to 110 different… Azure OpenAI Service Microsoft Sample Mar 11, 2024 Schneider Electric Schneider Electric is a digital partner for sustainability and efficiency, and its vast menu of connected devices, solutions, and services confirms this ambition. It builds objects as ubiquitous as circuit breakers for… Bonnard’s team laid the groundwork for Schneider’s EcoStruxure platform in 2009, well before Internet of Things… Azure Machine Learning · Azure OpenAI Service · Microsoft Cloud for Manufacturing 
