Nvidia case highlights Foxconn’s AI use in manufacturing
TechNews, citing a customer case published by Nvidia, reports that Foxconn used AI to optimize injection-molding parameter settings and changes, reducing the related engineering workload by 50%. Automated production scheduling reduced the hours required for manual scheduling by 25%, while AI-assisted SMT analysis addressed more than 20 major production-line defects, with an anomaly improvement rate above 90%. Nvidia also reports that Foxconn uses DGX SuperPOD and DGX GB200 systems to develop its FoxBrain language model, while Nvidia deployment and simulation tools support the rollout of Foxconn-developed AI technologies across factories.
The case details how Foxconn is applying AI to manufacturing parameters, production scheduling, defect analysis and model development, as well as using Nvidia infrastructure and tools across its operations.
鴻海持續深化與 Nvidia 的人工智慧合作，AI 技術導入製造現場的成果逐步浮現。據 Nvidia 最新公布的客戶案例，鴻海透過 AI 優化射出成型製程參數設定與切換，使相關工程工作量降低 50%；此外，導入 AI 自動排程後，排程工時也減少 25%，表面黏著技術（SMT）產線更改善超過 20 項主要缺陷問題，異常改善率逾 90%，展現 AI 從模型開發走向實際製造應用的效益。
Nvidia 指出，鴻海在全球擁有超過 200 座生產設施，業務涵蓋電子製造、電動車及智慧城市等領域。面對龐大的製造管理需求，鴻海透過導入 AI 技術，將原本仰賴工程人員經驗判斷及手動調整的作業，逐步轉向智慧分析與自動化決策，降低重複性工作所需的人力與時間。
#Foxconn#Nvidia#FoxBrain#DGX SuperPOD#DGX GB200#NVIDIA NeMo#NVIDIA Mission Control#NVIDIA Run#NIM#NVIDIA Omniverse#NVIDIA Isaac#VisionBay.ai