📄 中文摘要
建筑行业正面临生产力停滞、熟练劳动力短缺以及安全隐患等多重挑战。尽管机器人自动化技术为这些问题提供了潜在解决方案，但建筑机器人目前在适应非结构化、动态变化的施工现场方面仍面临困难。其核心症结在于即兴创作能力，即通过创造性问题解决来应对意外情况的能力，这在很大程度上仍是人类独有的特长。在建筑业这种高度不可预测的环境中，人机协作的即兴创作能力显得尤为重要，它能有效弥补机器人应对突发状况的不足，提升整体作业的灵活性和鲁棒性。为实现这一目标，需要构建一套系统性的理论框架，明确人机即兴创作的类型、层次及其在不同场景下的应用模式。这包括对即兴创作过程中的感知、决策、执行以及学习机制进行深入研究，尤其关注如何在机器人系统中融入类似人类的直觉和经验推理能力。同时，需要开发新的算法和交互范式，使机器人能够实时识别并理解非预期事件，并能够与人类操作员进行高效、自然的沟通，共同制定和执行即兴解决方案。研究路线图应涵盖传感器技术、机器学习、人机交互界面以及认知架构等多个交叉领域，旨在逐步提升机器人从被动响应到主动适应，乃至创造性解决问题的能力，最终实现人机在复杂建筑环境中无缝、高效的即兴协作。
📄 English Summary
Advancing Improvisation in Human-Robot Construction Collaboration: Taxonomy and Research Roadmap
The construction industry grapples with persistent productivity stagnation, a severe shortage of skilled labor, and significant safety concerns. While robotic automation offers promising avenues to address these issues, current construction robots struggle with adaptability to unstructured and dynamic site conditions. Central to this challenge is the capacity for improvisation, defined as the ability to creatively solve problems and adapt to unexpected situations, a faculty predominantly human. In the highly unpredictable environments characteristic of construction, collaborative human-robot improvisation is paramount for enhancing operational flexibility and robustness. To achieve this, a systematic theoretical framework is necessary to categorize the types and levels of human-robot improvisation and their application modes in diverse scenarios. This involves in-depth research into the perceptual, decisional, executive, and learning mechanisms underpinning improvisation, with a particular focus on integrating human-like intuition and experiential reasoning into robotic systems. Concurrently, novel algorithms and interaction paradigms must be developed to enable robots to recognize and comprehend unanticipated events in real-time, facilitating efficient and natural communication with human operators for joint formulation and execution of improvised solutions. The research roadmap should encompass interdisciplinary areas such as sensor technology, machine learning, human-robot interaction interfaces, and cognitive architectures. The ultimate goal is to progressively advance robots from passive responsiveness to active adaptation and creative problem-solving, thereby achieving seamless and highly effective human-robot improvisational collaboration in complex construction settings.