當工業 AI 重塑工廠與供應鏈之際，相關服務供應商的生態系統正迅速擴張。企業尋求運用 AI 來提升效率並加速創新，而新創公司則能創造價值，並從製藥、物流到航空航天等各領域帶來顛覆性變革。
歐洲的募資狀況也反映此趨勢：僅 2025 年第一季，歐洲 AI 公司便募得約 30 億歐元的創投資金，較 2024 年同期成長 55%。這股動能背後的原因很簡單：在企業必須面對極為嚴苛的時程與預算限制下營運的產業中，經營團隊迫切需要能重新構建工作流程、並大幅改善獲利表現的應用程式。
全新的生成式 [AI] 技術，能透過數位模擬與過去無法實現的超高效率，帶來嶄新的解決方案。
「這些技術不僅帶來效率上的躍進，部分更開創了全新的流程與工作模式。」WIRED Consulting 總經理 Catherine Crump 表示。本文收錄了包括她在內、來自日益壯大的 AI 生態系統中的多位專家的意見。
設計是 AI 能加速生產階段、導入全新效率的領域之一。
AI 工具讓工程師得以分析大量結構化與非結構化資料，透過模擬為替代材料與產品效能帶來全新視角，並協助解決長期存在的供應鏈瓶頸。
我預見未來的供應鏈將出現根本性重組，現有模式將受到巨大顛覆。
總部位於阿布達比、專注於醫療、航空航天及其他工業領域解決方案的 AI 研發公司 G42，其投資副總裁 Jon Nieman 表示，正是 AI 技術的融合與跨界應用，才促成了重大突破。
除了帶來顯著的獲利提升，AI 研發者更有機會在更大規模上創造正面影響。對新創顧問公司 embassidy 創辦人 Meike Neitz 而言，AI 在減緩氣候變遷上的潛力，是其最令人振奮的特點。舉例來說，透過設計階段的創新，能減少材料浪費——工業產品高達 80% 的環境足跡，早在設計階段便已決定。
結合積層製造、以 AI 重新設計的工業機器人夾爪，可讓單台機器人的碳排放量減少 82%。
「工業領域仍是全球溫室氣體排放的主要來源之一。」Neitz 表示。
AI 驅動的解決方案，能在提升能源效率、推動電氣化、減少材料浪費與優化流程上發揮龐大作用。
然而，即便對 AI 在現實世界的潛力抱持宏偉願景，也僅能走到這一步。正如本文專家所言，AI 新創公司創辦人必須具備清晰的領域知識、懂得擴張與協作，並且專注於最終使用者。
引領工業 AI 轉型，無法單靠一家公司獨力完成。這也是為什麼我們正打造蓬勃發展的工業 AI 生態系統，這需要客戶、產業領導者、新創公司、銷售團隊、合作夥伴與開發者之間緊密合作。
工業 AI 可輔助視覺模擬，為工廠現場帶來實質影響。
工業 AI 新創公司所面臨的挑戰之一，是訂立清晰的投資報酬 (ROI) 路徑，以及符合產業場景的應用情境。若無法做到這點，其應用程式很可能淹沒在眾多方案之中。
Crump 認為，新創公司必須致力解決產業中特定未被滿足的需求。抱持這樣的思維能建立競爭優勢。「當前 AI 領域充斥著炒作，也有大量新創運用這項技術進行研發。」她說，「能脫穎而出的，是那些能證明其獨特 AI 解決方案確實解決明確需求、並帶來更佳成果的團隊。其產品與服務能展現具體、正面的商業影響與清晰的投資報酬。」
新創公司創造影響力的方式，在不同工業場域中差異極大，這也反映出 AI 技術的多元應用潛力。新創公司必須深入理解其解決方案如何與工業硬體互動，或透過合作取得領域專業知識與高品質工業數據集。
Reimann Investors Venture Management 總經理 Samuel Schuler 建議，重點應放在事實、運用專屬數據，並提供可擴展、以使用者為中心的解決方案。「在客製化 AI 解決方案之前，應優先深入理解實際的工業工作流程。」他說。
量身打造、針對特定領域的洞察，成效往往優於通用型 AI 策略。
Alexander Oelling, the chief digital officer at ISAR Aerospace – a launch service provider for small and medium-sized satellites – also underscores the importance of real-world domain expertise.
I always try to assess founders' domain expertise beyond technology credentials.
"During discussions, I love to listen for unprompted mentions of specific challenges within target industries that wouldn't appear in superficial market research," says Oelling.
To ensure their solution delivers the promised impact on industry – and that its adoption spreads among users – startups need to demonstrate a laser focus on delivery. In practice, this means becoming extremely curious about the end user and how they engage with AI tools.
“Obsess about the people on the ground that you’re building your solution for,” embassidy’s Neitz advises founders and leaders. “Build for them, rather than their bosses. Be in touch with them to get their feedback. Learn their pains, about their work realities, about their processes.”
Developing user-centric AI is key to successful deployment.
Oelling also stresses the value of working closely with end-users, arguing that tangible, on-the-ground results are what cut through most when startups begin to establish themselves. “Industrial AI startups that secure early-pilot implementations, even if limited in scope or partially subsidized, establish credibility that theoretical ventures simply cannot match,” he argues.
Conversely, an insufficient grasp of the customer’s working reality can have a negative impact on sales, especially when targeting companies in utilities, oil and gas, aerospace and government, which have their own needs and preferences.
“I have seen it again and again,” says G42’s Nieman. “People fail to recognize that these types of customers do not want to dump their data into a data lake, they want two-year long pilots or vendor assessments, they tend to be fast-followers and they tend to pay less exciting multiples for startups."
Ultimately, this is about establishing credibility. The good news for startups that are successful in this endeavor is that they will find themselves in increasingly high demand.
As an investor, when trying to assess whether a startup will succeed, Nieman says he pays particular attention to its origin story. “I find the path is the best predictor of the future,” Nieman says. “Knowing the fabric of the inception of the company, the product, the technology and the team is more important than any one of those pieces in isolation. What helps a company stand out is understanding its roots – the skills, thinking and mindset of the founders, and the arc of commercialization of the product. All of this either verifies or unravels the story.”
The insights for this article were provided by the jury members of the Industrial AI Awards 2025 for startups at the AI with Purpose Summit.