TY  - JOUR
AU  - JIANG, Yong
PY  - 2026
TI  - Construction of philosophy and social science laboratories in industry-specific engineering universities driven by multi-agent collaboration
JO  - Experimental Technology and Management
SN  - 1002-4956
SP  - 279
EP  - 285
VL  - 43
IS  - 9
AB  - ObjectiveSince the Ministry of Education (MOE) launched the first batch of philosophy and social science laboratories in 2021, these laboratories have become important platforms for advancing interdisciplinary research and developing China's independent knowledge systems. However, engineering universities with industry-specific strengths in geology, mining, and petroleum face significant structural challenges in participating in this initiative. Their strong emphasis on the natural sciences and engineering has left philosophy and social science disciplines comparatively underdeveloped, even though the industries they serve present urgent interdisciplinary challenges requiring deep integration across the natural sciences, engineering, and the social sciences. This study addresses this gap by examining how such universities should develop philosophy and social science laboratories with distinctive industry characteristics, as well as the underlying mechanisms, evolutionary pathways, and institutional safeguards needed to support them.MethodsThis study introduces multi-agent system theory and incorporates large language models (LLMs) as both an analytical framework and an organizational design metaphor. It proposes a novel three-dimensional collaborative model involving discipline–technology–governance (DTG) agents. Discipline agents represent contributors from economics, management, and industry-specific engineering disciplines, each maintaining an independent knowledge base while participating in structured interdisciplinary knowledge exchange. Technology agents represent intelligent infrastructure that serves as a catalyst for bridging disciplinary boundaries through data fusion and methodological transfer. Governance agents represent institutional mechanisms responsible for coordinating stakeholder interests and monitoring organizational performance. These agents interact through a hybrid centralized–distributed architecture that combines top-down strategic coordination with bottom-up disciplinary self-organization. The framework is empirically validated through a case study of the MOE Social Science Laboratory of Mineral Resources Security Governance at China University of Geosciences (Beijing).ResultsThe study presents four major findings. First, it identifies four structural challenges: disciplinary ecosystem imbalance, reflected in the “strong engineering, weak humanities” contradiction in resource allocation; superficial interdisciplinary integration that remains episodic rather than institutionalized; a disconnect between technology and research, in which engineering platforms fail to extend beyond instrumental applications in social science research; and fragmented governance caused by departmental silos that hinder cross-unit collaboration. Second, the DTG model reveals three operational mechanisms: a knowledge production mechanism driven by interactions among heterogeneous agents, with technology agents serving as bridges across disciplinary boundaries; an organizational coordination mechanism that integrates micro-level self-organization with macro-level strategic oversight through a three-tier governance structure; and an evolutionary mechanism driven by both endogenous disciplinary needs and exogenous strategic, policy, and technological forces. Third, the case study demonstrates three distinctive characteristics of industry-embedded laboratory construction: an industry problem-oriented research approach anchored in sector-specific challenges; an engineering technology reverse-feeding mechanism that transforms technological strengths into social science data assets and methodological tools; and a government–industry–university–research integration pathway that establishes a closed-loop research ecosystem. Fourth, the study proposes a four-stage evolutionary model consisting of a disciplinary foundation stage for identifying anchor disciplines; a technology empowerment stage for developing an industry data hub and LLM-powered research assistant platforms; a collaborative integration stage for forming interdisciplinary teams through dual principal investigator arrangements and disciplinary bridge-builder programs; and an ecosystem construction stage for establishing multilayered open innovation networks that generate theoretical innovation, policy recommendations, and social service outcomes.ConclusionsThis study contributes to the field in three major ways. Theoretically, the DTG model moves beyond the prevailing paradigm of “technology instrumentalism” by reconceptualizing intelligent technologies as organizational mechanisms that reshape interdisciplinary knowledge production. Practically, the proposed four-stage evolutionary model demonstrates that the “strong engineering, weak humanities” profile of industry-specific engineering universities can serve as a latent comparative advantage rather than a limitation when compared with comprehensive universities. Empirically, the case of the MOE Social Science Laboratory of Mineral Resources Security Governance validates both the explanatory power and practical applicability of the DTG framework. The study recommends establishing differentiated evaluation standards for industry-specific universities, promoting sectoral data sharing between government agencies and laboratories, creating interdisciplinary talent development programs that include bridge-builder positions, and reforming evaluation systems to recognize originality, interdisciplinarity, and practical impact as core assessment criteria.
UR  - https://doi.org/10.16791/j.cnki.sjg.2026.09.035
DO  - 10.16791/j.cnki.sjg.2026.09.035