我們邀請到 DIGITIMES 副總經理黃逸平與 PowerArena CEO Ken，深入討論 AI 在智慧製造中的未來發展。黃副總分享了台灣在 AI 領域中的關鍵地位，Ken 則深入回應了製造業者在導入 AI 過程中，所面臨的實際挑戰與解決方案。兩位專家探討 AI 如何重塑生產流程，以及這些變革對產業發展的深遠影響，共同剖析在日益自動化、數位化的時代中，勞動力將如何轉型。

精彩時刻

- 00:00 片頭及來賓介紹

- 02:01 AI 世代的機會與挑戰

- 04:33 轉型升級：如何開始導入 AI 與避免障礙

- 12:39 正確的導入牽引指標

- 16:03 未來工廠：想像與現實

- 21:39 台灣的製造未來與供應鏈戰略

- 25:45 電動車前景：製造轉型優勢

PowerArena Website: https://www.powerarena.com/zh-hant/

LinkedIn: https://www.linkedin.com/company/powerarena

In this episode, we’re joined by Eric Huang, Vice President of DIGITIMES, and Ken, CEO of PowerArena, to discuss the future of AI in smart manufacturing. Eric shares insights into Taiwan’s role in the AI landscape, while Ken delves into the practical challenges and solutions manufacturers face when integrating AI. Together, they explore how AI is reshaping the production process and what lies ahead for the industry. Tune in to discover how AI is transforming manufacturing and what it means for the future of human workers in an increasingly automated world.

Highlights

- 00:00 Opening and guest introduction

- 02:01 Opportunities and challenges of the AI generation

- 04:33 Transformation and Upgrading: How to start introducing AI and avoid obstacles

- 12:39 Correct import of traction indicators

- 16:03 Factory of the future: imagination and reality

- 21:39 Taiwan’s manufacturing future and supply chain strategy

- 25:45 Prospects of electric vehicles: Advantages of manufacturing transformation

PowerArena Website: https://www.powerarena.com/

LinkedIn: https://www.linkedin.com/company/powerarena

PowerArena 資料科學家洪識濬 Mike 深入探討了大型語言模型 (LLM) 和 AI。他分析了近期釋出的 Claude 3.5 的表現，從使用者的角度解釋我們應該如何理解語言模型，並探討 AI 中無法避免的偏見問題。Mike 也分享了他對於個人在這個快速變化的 AI 時代，如何自我適應、順應成長。

精彩時刻

-00:32 Antropic Claude 3.5 表現怎麼理解？

-06:23 使用者視角下的語言模型運用

-08:26 為什麽語言模型必定含有偏見？Constitutional AI（合憲 AI）是什麼？有幫助嗎？

-16:31 AI 浪潮下的個體應對之道

PowerArena Website: https://www.powerarena.com/zh-hant/

LinkedIn: https://www.linkedin.com/company/powerarena

PowerArena’s Data Scientist, Mike Hung, dives into the fascinating world of LLM (large language models) and AI. He discusses the performance of the recently released Claude 3.5, how we should understand language models from a user’s perspective, and explores the inevitable biases in AI systems. Mike also offers his insights on how individuals should navigate this rapidly evolving AI landscape.

Highlights

-00:32 How to understand the performance of Anthropic Claude 3.5?

-06:23 LLM applications from a user's perspective

-08:26 Why are language models inherently biased? What is Constitutional AI, and does it help?

-16:31 How individuals can adapt in the AI era?

PowerArena Website: https://www.powerarena.com/

LinkedIn: https://www.linkedin.com/company/powerarena

#llm #claude3.5 #anthropic #openai #constitutionalai #ai #aivision #smartmanufacturing #industry40

在本集內容中，邁特副總經理＆邁特創新基地執行長 戴憶帆 Ray 向我們分享了邁特如何從 30 年老字號的電子 OEM／ODM 廠轉型，成為協助多家新創在製造路上的成長夥伴。在地緣政治與缺工的雙重挑戰之下， Mighty Net 分享了企業在轉型過程中的 Top-down 思維，為公司轉型建立緩衝，並且結合完善的基礎設施，讓新技術的導入更加順利，從而創造企業新的成長曲線。

精彩時刻

00:00-02:00 30 年老字號邁特電子如何轉型成創新推手？

02:00-06:30 從客戶到夥伴，邁特電子創新生態系

06:50-10:00 智造趨勢：人因分析與瑕疵檢測

10:00-13:09 Top-down 克服導入陣痛？

13:09-15:03 如何建立研發團隊的創新基因？

15:03-17:57 規模擴張：建立下一世代的基礎建設

17:57-23:08 70/30法則：設置轉型緩衝

23:08-28:10 拋開過去糖衣：台廠製造代工到全球技術輸出

In this episode, Ray Dai, VP of Might Electronics and CEO of Mighty Net, shares how Might transformed from a 30-year-old electronic OEM/ODM manufacturer into a growth partner for numerous startups. Amidst the dual challenges of geopolitical issues and labor shortages, Mighty Net discusses their top-down approach to corporate transformation, which established a buffer for the transition and leveraged robust infrastructure to facilitate the smooth integration of new technologies, thereby creating a new growth trajectory for the company.

Highlights

00:00-02:00 How did Mighty Electronics transform into a new venture catalyst?

02:00-06:30 From customers to partners, the innovative ecosystem of Mighty Net

06:50-10:00 Manufacturing Trends - Human Factor Analysis and Defect Detection

10:00-13:09 How to overcome the pains of adoption? Top-down is the key!

13:09-15:03 Establishing an innovation gene, driving industry upgrades with the R&D team

15:03-17:57 Scale expansion: Establishing infrastructure for the next generation

17:57-23:08 Embracing the 70/30 rule, establishing a buffer between transformation and stability.

23:08-28:10 Seize the opportunity to transform and upgrade, moving towards technological exports.

「AI 除了偵錯外，還可以辨認瑕疵種類進而分類，判別流程中是哪一段出了問題」

本集邀請到研華科技事業發展經理 紀明傑 Ray Chi 分享 AI 智造經驗，從最根本的思考架構開始，Ray 闡述如何運用 AI 思維，應對產業變遷，以工廠導入 AI 的實例，引領我們一探智造產線樣貌，掌握 AI 驅動的智慧製造脈動。

精彩時刻

00:00 intro

- 智造現狀＆案例解析

01:21 全球產業變動，AI 思維重整產線

06:36 AI 視覺主動偵錯，人力管理升級

10:50 AI 應用零死角：產能、良率、工安

- AI 落地 101

14:35 跳出框架分析：外部團隊打造適配軟硬體

19:51 AI 導入困境：OT＆IT 的專業整合

- AI 產業發展趨勢

22:12 AI 智造未來：邊緣 AI 伺服器

"Beyond error detection, AI possesses the capability to identify and categorize various defect types, pinpointing exactly where in the process issues arise."

In this episode, we're joined by Ray Chi, Business Development Manager at Advantech, who shares his insights into leveraging AI for smart manufacturing. Starting from the foundational framework, Ray delves into how AI thinking can be applied to adapt to shifts in the industry. Through real-world examples of AI integration in factories, he guides us through the landscape of smart manufacturing lines, helping us to capture the essence of AI-driven smart manufacturing advancements.

Highlights

00:00 intro

- Smart Manufacturing & Case Analysis

01:50 Global Shifts: AI Reshapes Production Lines

06:14 AI Vision Error Detection: Upgrade Traditional Management

10:32 AI Agility: Application in Capacity, Yield & Safety

- AI Implementation 101

14:35 Thinking Outside the Box: Utilizing External Expertise for Tailored Hardware and Software Solutions

19:37 Challenges in AI Implementation: Integration of OT & IT

- AI Industry Trends

22:20 Future of Manufacturing: Edge AI

本集 PowerTalk 很榮幸邀請到台達電企業策略業務發展和聯盟技術經理 梁哲銘 Kevin。Kevin 擁有與全球大型客戶合作的豐富經驗，熟悉領先大廠的目光焦點與未來策略，為頗析產業動態，帶來不同方向與高度，在「智造」趨勢上提供深入的見解。

轉型並非一蹴可及，瞭解台達電如何穩健地步步推進，為智慧製造扎根，也為永續打造多元解方，為企業創造更多價值。

This time on PowerTalk, we are privileged to welcome Kevin Liang, Delta Electronics, Inc, Corporate Strategic Business Development and Alliance, Technical Account Manager. With his extensive experience in working with major global clients and a keen understanding of industry leaders’ strategies and future roadmaps, Kevin offers his unique and in-depth insights of smart manufacturing.

Transformation is a marathon for every manufacturer, not a sprint. Find out how Delta steadily lays a strong foundation for smart manufacturing, and crafts diverse solutions for sustainability. Discover now how to create values for your business. Learning from the pioneer.

強化全球製造：AI 可視化海外生產管理

PowerArena 共同創辦人兼 COO Hauman 從第一線協助全球製造大廠的經驗，分享 AI 視覺如何導入東南亞產線，協助穩定海外生產基地。

為進行產線遷移、擴大產能的新基地建設，全球製造商們將眼光投放於東南亞、印度和墨西哥等人力資源充裕，且鄰近龐大市場的地區。但這類勞力密集產線的有個共同難題：如何有效提升產能？

面對龐大基數的作業員，過往的生產經驗有辦法複製到新廠嗎？ＡI 視覺對異地生產的價值在哪裡？

AI Visualization In Remote Factory Management

PowerArena co-founder and COO, Hauman, shares his frontline experience assisting a global top 5 manufacturer to implement AI vision in their factory in SEA to stabilize overseas production.

In an effort to transfer production bases and expand capacity, global manufacturers focus on regions like SEA, India, and Mexico where there are abundant human resources and proximity to vast markets. However, these labor-intensive production lines share a common problem: a difficulty in effectively increasing productivity.

Will past experiences from mature production lines be able to replicated in new factories? How could manufacturers manage this large number of workers? What values could AI vision bring to these remote factories?

想像vs.現實：AI智造到底意味什麼？

產線管理者期待 AI 能「看」到的資訊和現實有什麼落差？在追求 AI 智造前，應具備哪些對人工智慧的必要認知？

工站攝影機捕捉的影像，即是 AI 學習辨認產線的基礎，也是而後 AI 發揮分析動能之處。

AI 總監 Kuei 破除迷思，剖析 AI 智造的關鍵核心。

AI In Smart Manufacturing. What's Next?

Video footage captured by a camera installed on a workstation serves as the foundation for AI to learn and recognize the production line. That’s where the journey starts. But how can AI help manufacturers from there? And what is AI, exactly?

AI Director Kuei answers the long-due questions and breaks down the core of AI applications for smart manufacturers.

幕後揭秘 Gogoro 的 AI 佈建

Gogoro 積極地尋求品管解方，小至每一個生產節點，都希望完整掌握，以便維持優良產品品質、最大化產線效能。

從初期站點評估、MES 系統整合到 AI 的成功運用，PowerArena 專案經理 Ian 解說專案執行過程中的考量與挑戰，一步步同 Gogoro 建構串連 IoT、AGV 的 AI 視覺系統，打造出真正的智慧工廠。

Behind The Scenes of Gogoro's AI Deployment

Gogoro is actively seeking smart solutions for IPQC, even for the smallest production detail.

PowerArena’s project manager, Ian, shares his considerations during the Gogoro project and the challenges lied along the way of AI implementation. From the initial assessment, MES system integration, all the way to the success of AI deployment, Ian shares his first-hand experience.

Listen how Gogoro worked with PowerArena to build an AI vision system that connected IoT and AGV to truly turn the concept of smart factory into reality.

AI 解構 IE 工程師 Domain Know-how 數位化

投身製造業三十餘年的呂俊毅廠長 Junny 不藏私地分享工廠的管理訣竅。

在工廠面臨的種種挑戰中，為什麼「人」是最不易管理的因素？

大家一直追求減少成本、提升產值，效益卻有限？

數位轉型浪潮裡，工廠紛紛希望晉身智慧工廠，而制勝的關鍵便是使用對的工具，簡化繁瑣的舊日常，跳脫既有思維，全面升級管理。

Junny 直白的解說將會為我們帶來更多的新想法！

An AI deconstruction of IE

Junny Lu, who has been in the manufacturing industry for nearly 30 years, is sharing his tips on smart manufacturing.

Among the 4M1E, why is "man" the most difficult factor to manage? Try hard to reduce costs and increase output but can't see significant results?

In the wave of digital transformation, factories are eager to transform into smart factories. The key to standing out is to use the right tools and make them our supercharger while getting rid of old mindsets at the same time.

Gogoro 智慧工廠導入 AI 視覺

第三集 Powertalk 邀請 Gogoro 智慧製造系統及解決方案 處長 黃至偉 揭秘機車製造領頭羊的工廠秘辛，分享如何突破管理限制，將已經部署多樣智慧工具如 IoT、AGV 等的智慧工廠，再次升級。

借助 AI 視覺之力，數據化原先難以捕捉的「人員動作」及「無法聯網的設備使用」，全面排除品質疑慮，真正實現製程品管。第一線管理專家詳細揭示 Gogoro 的智慧製造佈建。

Gogoro Smart Factory Implements AI Vision

PowerTalk invites Charlie Huang, Manager of Smart Factory Platform & Solutions at Gogoro, to reveal the secrets of how the leading E-scooter manufacturer upgrades their smart factories. Already has deployed multiple smart tools, like IoT and AGV, Gogoro breaks through the limitations of management and takes their smart factories to the next level with the assistance of AI.

“Human motions” and “offline equipments operations” can now be digitized with AI vision. IPQC is no longer a concept that lacks complete feasibility, but a real action that could bring immense benefit to the production. Listen to the first-line expert disclose Gogoro's smart manufacturing deployment.