AI News AI资讯 4h ago Updated 1h ago 更新于 1小时前 45

Qisu Technology, a Zhejiang University-affiliated Desktop CNC Team, Raises Nearly 100 Million RMB in Angel Round from SenseTime Guoxiang, Shouxing Technology, etc., to Lower Manufacturing Barriers with AI 36氪首发 | 浙大系桌面CNC团队获商汤国香、首形科技等近亿元天使轮,要用AI技术降低制造门槛

Qisu Technology secured nearly 100 million RMB in angel funding from SenseTime Guoxiang, Shouxing Technology, and others to develop desktop CNC machines powered by AI. The company addresses the high barrier to entry in CNC manufacturing by developing an "AI CAM" system that automates complex process planning, eliminating the need for manual parameter configuration. Core technical challenges involved digitizing tacit knowledge from experienced operators and handling vast real-world variables, sol 浙大系团队创立的桌面级CNC企业「奇塑科技」完成近亿元天使轮融资,投资方包括商汤国香、首形科技等。 公司核心突破在于自研AI CAM系统,旨在解决传统CNC操作门槛高、依赖专家经验的行业痛点。 通过自建柔性工厂积累数十万份真实加工数据,构建数据闭环以训练可复用的工艺决策模型。 产品交互逻辑反转,用户仅需导入模型即可自动生成工艺,实现“端到端”易用性,计划Q4发布首款桌面五轴CNC。 创始人强调长期主义,指出AI CAM需经过数年真实场景验证才能达到规模化落地的稳定性标准。

65
Hot 热度
60
Quality 质量
65
Impact 影响力

Analysis 深度分析

TL;DR

  • Qisu Technology secured nearly 100 million RMB in angel funding from SenseTime Guoxiang, Shouxing Technology, and others to develop desktop CNC machines powered by AI.
  • The company addresses the high barrier to entry in CNC manufacturing by developing an "AI CAM" system that automates complex process planning, eliminating the need for manual parameter configuration.
  • Core technical challenges involved digitizing tacit knowledge from experienced operators and handling vast real-world variables, solved through a closed-loop data system built via their own flexible factory.
  • The product strategy shifts interaction logic from "user adapts to software" to "software adapts to user," offering a three-step workflow: process generation, preview, and machining.
  • A flagship desktop five-axis CNC machine with automatic tool changing and a fully enclosed cabin is scheduled for release in Q4 2026, targeting the consumer and prosumer markets.

Why It Matters

This development signals a critical shift in the democratization of advanced manufacturing, moving beyond additive technologies like 3D printing to subtractive methods capable of processing hard materials. By leveraging AI to automate Computer-Aided Manufacturing (CAM), Qisu Technology aims to solve a decades-old industrial problem, potentially unlocking a trillion-dollar market in consumer-grade precision manufacturing. For the industry, it highlights the growing convergence of industrial automation, edge computing, and accessible hardware, suggesting that professional-grade capabilities may soon become standard for hobbyists and small businesses.

Technical Details

  • AI CAM Architecture: The core innovation is an end-to-end AI decision-making model that replaces traditional manual CAM programming. It ingests 3D models and automatically generates toolpaths, selecting tools, speeds, and feeds without human intervention.
  • Data-Centric Approach: To overcome the lack of digital assets in traditional CNC, the team built a flexible factory to collect hundreds of thousands of real-world machining cases. This data loop allows the AI to learn from "corner cases" and continuously refine its predictive accuracy for diverse materials and geometries.
  • Hardware Specifications: The upcoming desktop five-axis CNC features a 1500W high-power spindle, automatic tool changer (ATC), automatic tool setting, and a fully enclosed dust-proof and silent cabin, designed for safe home or office use.
  • User Interface Simplification: The software interface reduces the complex multi-parameter setup of traditional CAM into three intuitive steps: import model, generate process, and execute, effectively hiding the underlying computational complexity from the user.

Industry Insight

  • Market Expansion Beyond 3D Printing: As the consumer 3D printing market saturates, the next wave of growth lies in hybrid manufacturing devices that can handle both soft and hard materials. Investors and manufacturers should watch for hardware-software bundles that lower the skill floor for subtractive manufacturing.
  • The Value of Proprietary Data Moats: In industrial AI, algorithmic superiority is often secondary to data quality. Companies that build physical infrastructure to generate proprietary, real-world manufacturing data will hold significant competitive advantages over those relying solely on synthetic or public datasets.
  • Long-Term R&D in Deep Tech: The timeline mentioned (algorithm stability taking years to achieve) underscores that AI in heavy industry requires long-term capital patience. Short-term hype cycles may fail here; sustainable success depends on continuous iteration in real-world environments rather than quick software updates.

TL;DR

  • 浙大系团队创立的桌面级CNC企业「奇塑科技」完成近亿元天使轮融资,投资方包括商汤国香、首形科技等。
  • 公司核心突破在于自研AI CAM系统,旨在解决传统CNC操作门槛高、依赖专家经验的行业痛点。
  • 通过自建柔性工厂积累数十万份真实加工数据,构建数据闭环以训练可复用的工艺决策模型。
  • 产品交互逻辑反转,用户仅需导入模型即可自动生成工艺,实现“端到端”易用性,计划Q4发布首款桌面五轴CNC。
  • 创始人强调长期主义,指出AI CAM需经过数年真实场景验证才能达到规模化落地的稳定性标准。

为什么值得看

本文揭示了AI技术在传统制造业(特别是工业软件CAM领域)落地的具体路径与难点,展示了如何通过数据闭环解决非结构化工艺知识的数字化难题。对于关注AI+制造、工业软件国产化及消费级硬件创新的从业者而言,提供了关于如何将专家经验转化为算法模型的实战参考。

技术解析

  • AI CAM核心架构:摒弃传统“人适应软件”的模式,采用“软件适配人”的交互逻辑。系统通过AI算法自动处理复杂的数控编程,将工艺生成、预览、加工收敛为极简三步,无需用户掌握专业加工知识或手动配置上百项参数。
  • 数据驱动的知识沉淀:针对传统CNC工艺知识依赖“口传心授”、难以数字化的问题,团队通过自建基于AI CAM的柔性工厂进行真实接单生产,积累数千小时实操经验和数十万份真实加工模型数据,构建数据闭环以持续迭代算法。
  • 长周期算法验证:强调AI CAM研发的长期性,指出即使资深工程师也需半年以上精通,而算法稳定性需经长期验证。例如某核心算法从2021年编写至2024年底才满足规模化落地要求,体现了对Corner Case(边缘案例)的高覆盖率处理需求。
  • 硬件集成规格:即将发布的桌面级五轴CNC标配自动换刀(刀库)、自动对刀、全封闭防尘静音舱体及1500W大功率主轴,结合AI CAM系统实现从软件到硬件的端到端易用性。

行业启示

  • 工业软件的AI重构机遇:CAM作为制造业核心软件,其自动化程度低、门槛高是长期痛点。利用AI将隐性专家知识显性化、模型化,是打破工业软件壁垒、实现普惠制造的关键突破口。
  • 数据闭环是AI落地的核心壁垒:纯算法难以解决复杂物理世界的加工问题,必须结合真实生产场景构建“数据-算法-验证”闭环。自建柔性工厂或类似实体场景是获取高质量、多样化工艺数据的必要手段。
  • 消费级制造设备的下沉趋势:随着3D打印等技术的市场教育完成,具备更强材料适应性的桌面级CNC有望成为下一个万亿级蓝海。降低使用门槛(易用性)是此类设备能否从小众极客走向大众消费者的决定性因素。

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Funding 融资 Robotics 机器人 Product Launch 产品发布