建筑业是国民经济支柱产业。由于人口老龄化、新冠疫情及施工环境恶劣等多重因素影响，建筑业面临日益严峻的劳动力短缺挑战。考虑到施工环境的复杂性和动态性，研发全自动机器人仍面临诸多挑战。因此，未来很长一段时间内，工人与机器人将在工程现场并存并相互协作，以高效地建造或维护设施。当前，人机协同（HRC）建造作为一个新兴领域，仍然面临着各种开放性问题。针对人机协同问题，本研究开创性研究引入了一种基于代理的建模方法来研究 HRC在砌砖过程中的耦合效应和规模/尺度效应。多次仿真实验表明，HRC 的动态和复杂性体现在两个方面：1）HRC 中的智能体由于工人的人因特征、机器人的性态参数及人机协同行为影响而相互依赖； 2）HRC的不同参数相互关联依赖，对施工效率（CP）具有显著影响。同时研究表明，HRC 对 CP 具有明显的规模效应或尺度效应，这意味着即使人机比例保持不变，增加人机协作团队的数量也会明显提升CP。研究同时建议，未来可围绕施工效率提升、工人职业安全等对HRC进行更多研究。本研究可以为开发和评估新型机器人、优化人机协同流程以及培训未来产业工人等奠定基础。
As an important contributor to GDP growth, the construction industry is suffering from labor shortage due to population ageing, COVID-19 pandemic, and harsh environments. Considering the complexity and dynamics of construction environment, it is still challenging to develop fully automated robots. For a long time in the future, workers and robots will coexist and collaborate with each other to build or maintain a facility efficiently. As an emerging field, human-robot collaboration (HRC) still faces various open problems. To this end, this pioneer research introduces an agent-based modeling approach to investigate the coupling effect and scale effect of HRC in the bricklaying process. With multiple experiments based on simulation, the dynamic and complex nature of HRC is illustrated in two folds: 1) agents in HRC are interdependent due to human factors of workers, features of robots, and their collaboration behaviors; 2) different parameters of HRC are correlated and have significant impacts on construction productivity (CP). Accidentally and interestingly, it is discovered that HRC has a scale effect on CP, which means increasing the number of collaborated human-robot teams will lead to higher CP even if the human-robot ratio keeps unchanged. Overall, it is argued that more investigations in HRC are needed for efficient construction, occupational safety, etc.; and this research can be taken as a stepstone for developing and evaluating new robots, optimizing HRC processes, and even training future industrial workers in the construction industry.
This work was supported by National Natural Science Foundation of China (No. 51908323, No. 72091512) and the National Key R&D Program of China (No. 2018YFD1100900).