Scholarlink

<p>3D printed concrete (3DPC) technology is driving the construction industry toward automation and sustainable practices. However, its widespread adoption remains hindered by inherent material anisotropy, unpredictable process control, and complex structural design. To overcome these bottlenecks, artificial intelligence (AI) has emerged as a transformative solution. This paper provides a comprehensive review of AI in 3DPC across three core dimensions, highlighting a paradigm shift from an empirical, open-loop pipeline to a unified cyber-physical framework driven by bidirectional information feedback. At the material level, machine learning (ML) enables inverse design and multi-objective optimization of mix proportions. During the printing process, the integration of machine vision and adaptive control establishes a robust perception-decision-execution closed-loop system, ensuring deposition quality and geometric fidelity. At the structural level, generative design and topology optimization facilitate the creation of complex geometries. Meanwhile, AI models enable multi-scale performance evaluations, ranging from micro-defect identification to macro-scale load-bearing capacity assessment. Despite these achievements, bottlenecks such as data heterogeneity and physics-agnostic models persist. Future research is expected to focus on cross-layer coupling, physics-informed modeling, and digital twin interoperability. By continuously feeding process execution and structural evaluation data back into material formulation, this new paradigm is poised to transform 3DPC into a fully self-adaptive and autonomous construction ecosystem.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AI-powered 3D printed concrete technology: a new paradigm of intelligent construction integrating materials, equipment, and structures

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月8日

查看全文
  • Li Wang,
  • Zupan Zhang,
  • Yupeng Guo,
  • Jinggang Xu,
  • Yahong Zhou

摘要

3D printed concrete (3DPC) technology is driving the construction industry toward automation and sustainable practices. However, its widespread adoption remains hindered by inherent material anisotropy, unpredictable process control, and complex structural design. To overcome these bottlenecks, artificial intelligence (AI) has emerged as a transformative solution. This paper provides a comprehensive review of AI in 3DPC across three core dimensions, highlighting a paradigm shift from an empirical, open-loop pipeline to a unified cyber-physical framework driven by bidirectional information feedback. At the material level, machine learning (ML) enables inverse design and multi-objective optimization of mix proportions. During the printing process, the integration of machine vision and adaptive control establishes a robust perception-decision-execution closed-loop system, ensuring deposition quality and geometric fidelity. At the structural level, generative design and topology optimization facilitate the creation of complex geometries. Meanwhile, AI models enable multi-scale performance evaluations, ranging from micro-defect identification to macro-scale load-bearing capacity assessment. Despite these achievements, bottlenecks such as data heterogeneity and physics-agnostic models persist. Future research is expected to focus on cross-layer coupling, physics-informed modeling, and digital twin interoperability. By continuously feeding process execution and structural evaluation data back into material formulation, this new paradigm is poised to transform 3DPC into a fully self-adaptive and autonomous construction ecosystem.