智能制造是推动我国制造业转型升级、提升产业竞争力的系统工程，也是实施制造强国战略的关键环节，深入推进智能制造需要坚持思想创新，从哲学根源出发制定科学发展战略。本文采用现象学还原论和本体论的透视方法，从技术哲学视角出发，集成系统辩证逻辑推理，识别智能制造人机协同技术“难问题”，基于科学知识论的立场洞察相应哲学本质。研究发现，显性知识和隐性知识普遍存在于制造技术活动中，而长期以来的技术发展忽略了隐性知识的重要性；对人类主体存在不合理的“理性人”假设，忽视了其在制造系统中发挥的关键作用。为此，通过本体技术反思，提出了基于“物联网、知联网、身联网”的笛卡尔式智能制造技术发展路径，进而建立了基于“行为导向、指示表征、具身嵌入”的海德格尔式智能制造系统架构；由此溯源推因，构建了面向“难问题”消解的“数据驱动、功能表征、具身融合”三阶段递进发展的人机协同知识范式。为了高效推动新一代智能制造发展，可构建知识工程分类评估体系，应用理论认知模型来推动技术难题解决并制定产业扶持政策，建立多侧应用型智能制造公共服务平台以强化知识流通与整合应用。

Intelligent manufacturing is a systematic engineering and technological innovation that propels the transformation and upgrading of China’s manufacturing industry, enhancing industrial competitiveness. It is a pivotal element in strengthening the manufacturing sector of China. The comprehensive and profound advancement of intelligent manufacturing necessitates adherence to ideological innovation, requiring the formulation of a scientifically grounded development strategy from its philosophical roots. This paper, based on the phenomenological reductionism and ontological perspective method, starts from the perspective of the philosophy of technology, integrates systemic dialectical logical reasoning, investigates and identifies the “hard problems” of human– machine collaborative technology in intelligent manufacturing, and insightfully explores its philosophical essence from the standpoint of the philosophy of scientific epistemology. The research findings indicate that explicit knowledge and tacit knowledge are commonly present in manufacturing technology activities. However, the long-standing technological development has overlooked the significance of tacit knowledge and has made an unreasonable assumption about the existence of “rational people” in human subjects, neglecting their crucial role in manufacturing systems. Therefore, this paper, through ontological reflection, proposes a Cartesian intelligent manufacturing technology development path based on the Internet of Things, Internet of Contents & Knowledge, and Internet of Bodies. Building upon this foundation, it establishes a Heideggerian intelligent manufacturing system architecture grounded in behavior-oriented, deictic representations, and embodied embedding. Through this traceable causal analysis, it constructs a three-stage progressive development knowledge paradigm for resolving the “hard problems” of human – machine collaborative technology. This paradigm is characterized by data-driven, functional representation, and embodied integration. To efficiently promote the development of the new-generation intelligent manufacturing, a knowledge engineering classification assessment system can be established. The application of theoretical cognitive models can drive the resolution of technological challenges and the formulation of industry support policies. Establishing a multi-faceted application-oriented intelligent manufacturing public service platform strengthens knowledge circulation and integrated application.