📌 一句话摘要 对话西门子及工业界专家，深度拆解工业 AI 落地中关于确定性、数据孤岛及物理 AI 融合的核心挑战与实战经验。 📝 详细摘要 本期播客记录了西门子 RXD 大会上的工业 AI 圆桌论坛。对话嘉宾包括西门子中国研究院院长朱骁洵、成都数字化工厂厂长李永利及深度智控创始人李辉。内容深入探讨了工业 AI 与互联网 AI 的本质区别：前者追求极致的「确定性、安全性和可靠性」，而非单纯的「满意度」。专家们分析了工业 AI 落地的核心障碍，包括 IT 与 OT 系统的数据孤岛、高质量数据集的匮乏以及从标准方案到现场适配的「最后一公里」难题。讨论还重点提及了物理 AI 的前沿趋势，强调 A
Title: 68. [Extra] Industrial AI Pitfall Guide: When 180 Years o...
URL Source: https://www.bestblogs.dev/podcast/6d7c8d4
Published Time: 2026-03-27 01:26:01
Markdown Content:
Skip to main content
Toggle navigation menu Toggle navigation menuArticlesPodcastsVideosTweetsSourcesNewsletters
⌘K
Change language Switch ThemeSign In
Narrow Mode
Sign in to bookmark podcasts and track your listening history. Sign in now
68.【番外】工业 AI 避坑指南：当 180 年工业技术积淀，遇到大模型
One Sentence Summary
A conversation with Siemens and industry experts, deeply dissecting the core challenges and practical experiences in implementing Industrial AI, focusing on determinism, data silos, and the integration of Physical AI.
Podcast Info
From:卫诗婕｜商业漫谈Jane's talk
Published At:Today
AI Score:
89
Categories
Business & Tech
Chinese
#### Tags
Industrial AI Siemens Physical AI Digital Transformation Generative AI
Your browser does not support the audio player.
!Image 4: Podcast cover
卫诗婕｜商业漫谈Jane's talk
Now Playing
68. [Extra] Industrial AI Pitfall Guide: When 180 Years of Industrial Expertise Meets Large Models
00:00 00:00
1x
Show Notes Content Overview Transcript Q&A Recap
!Image 5
3月23日，我受邀主持了西门子RXD大会的工业 AI 圆桌论坛。
西门子在工业 AI 赛道沉淀了 50 年，作为全球领军企业，面对这一轮 AI 浪潮，有着自己的理解和布局。
当下，AI 浪潮正席卷全球各个行业，但真正落到工业场景里，却常面临着「落地难、规模化更难」的困境。我在现场，和三位工业界的一线资深专家，深度拆解了工业 AI 落地的核心挑战和应对思路。
中国是工业大国，AI在这个领域的实践，将书写一场不同的叙事。这期节目，分别从技术架构、工厂实战、跨界观察三个维度，还原了工业 AI 的真实落地图景。
三位嘉宾分别是：
朱骁洵（西门子中国研究院院长、工业 AI 顶层战略与技术研发权威）
李永利（西门子成都数字化工厂厂长、工业 AI 落地一线操盘手）
李辉博士（深度智控创始人、物理 AI 领域实战专家）
> 本期节目由西门子特别赞助。
本期Shownotes：
05:35 工业界对于AI的焦虑：期待高、落地难
06:58 工业AI 改造车间，生成式AI提效办公室工作
10:18 人机交互只是开始，AI有机会解决许多长尾需求
12:24 制造业的数据难题：数据孤岛、IT与OT系统的自动化难题
13:37 企业如何把数据收集起来，注重数据的可连接性和质量
17:15 这一轮 AI，没有来得及做好数字化的公司，还有机会吗？（嘉宾辩论了～😁）
> 工业场景数据非常稀缺
> 没有质量的数据带不来好模型
> 新工具有机会加速一切
20:46 多模态能力的提升，将提升模型在虚拟与物理世界的互动
22:47 硬件是数据的入口，数据是模型的燃料
23:21 谈到工业 AI ，一定要聚焦场景，「最后一公里」
27:59 不要总算人效，关注长期能力的建设
34:19 AI在工厂对人的替代：职工转向高级工种，收入也更多
36:44 将所有显性知识，灌进智能体
40:15 工业场景仍要注重，将工作流模块化
40:46 「让一线员工，坐在副驾驶」
44:55「 物理 AI 不解决工业所有问题」
48:01 建模、理解、学习、控制
加入听友群👇
!Image 6
Related Podcasts
阿里云开发者 ### Building Thinking Test Agents: The Evolution from Automation to Autonomous Intelligence Alibaba Cloud shares how to build 'Quality Digital Humans' with self-awareness through LLMs and multi-agent architectures, achieving an evolution from passive automation to proactive intelligent testing.腾讯科技 ### After Reading 200 Papers: Analyzing DeepMind, Meta, DeepSeek - What AGI Narratives Are Chinese and American Giants Describing? | 2025 AI Annual Review This article reviews 200 AI papers from 2025, deeply analyzing breakthroughs and bottlenecks in AGI across four key areas: reasoning, memory, spatial intelligence, and meta-learning, while looking ahead to 2026 development trends.钉钉 ### Full Transcript | DingTalk CEO Wu Zhao: Breaking Down DingTalk, Rebuilding with AI, and Forging 'Wukong' DingTalk CEO Wu Zhao launched the AI-native work platform 'Wukong,' proposing a shift in the AI interaction paradigm from GUI to CLI, and enabling autonomous operation and security management of enterprise software through DingTalk CLI and the RealDoc system.51CTO技术栈 ### Major Update! Jensen Huang's First Keynote of the Year: The World is Undergoing Two Platform Revolutions! New Rubin Computing Architecture Debuts! First Autonomous Driving AI Open-Sourced, AI-Native Application Framework Released! Grok 5 Parameter Scale Revealed At the CES 2026 opening keynote, Jensen Huang analyzed the two platform revolutions in AI-driven computing and applications, unveiling the new Rubin computing architecture, Nemotron 3 open-source models, a physical AI system based on Cosmos and Alpamayo, and the AI-native application framework Blueprint, while emphasizing NVIDIA's full-stack AI strategy and open ecosystem.卫诗婕｜商业漫谈Jane's talk ### 62. Year-End Review AI Notes: What Insights Emerged from Interviews with Over 30 Practitioners in China and the US? This episode provides a comprehensive review of the collisions and integrations in the 2025 AI industry between China and the US, covering technological leaps, geopolitical competition, capital flows, and business model exploration.卫诗婕｜商业漫谈Jane's talk ### 60. From Evolutionary Theory to the Evolution of Algorithms: The Birth of a New Paradigm as Large Models Aid Algorithm 'Reproduction' This podcast delves into how Baidu's 'FaMou' product leverages Large Language Models to drive algorithmic self-evolution, reshaping industrial productivity and production relations, and analyzes its innovation in cutting-edge exploration and commercial implementation across China and the US.卫诗婕｜商业漫谈Jane's talk ### 61. Minimax Goes Public: Three Years of Evolution in the LLM Landscape | A Conversation with Chen Yu from Yunqi Partners This episode explores Minimax's IPO, angel investor Chen Yu's investment philosophy, and the three-year evolution and future outlook of China's LLM industry across technology, business, and competition.卫诗婕｜商业漫谈Jane's talk ### 63. Interview with Rokid's Misa Zhu: AI Glasses, Entry Points, and the Ecosystem Battle Amid Tech Giants Rokid founder Misa Zhu provides an in-depth analysis of the growth logic, competitive landscape, and ecosystem strategy of AI glasses as the next generation of spatial computing terminals.
HomeArticlesPodcastsVideosTweets
68. [Extra] Industrial AI Pitfall Guide: When 180 Years o...