Enhancing Construction Safety through Hazard-Aware Human-Robot Task Allocation

Construction robots can free workers from dangerous and tedious tasks, but unstructured construction environments make it difficult for a single robot to perform complex tasks efficiently. Although human-robot collaboration (HRC) is increasingly needed, frequent interactions between humans and robots inevitably pose unprecedented safety concerns, requiring hazard-aware human-robot task allocation. To this end, this paper proposes a new task allocation method for HRC. First, a decomposition method for construction tasks is introduced, and the criteria for judging the attributes of HRC are established. Second, the hazardous events construction workers face in HRC scenarios are analyzed. Third, a multicriteria human-robot construction task allocation (HRCTA) model is developed with the hazard level as the main objective. The feasibility of the HRCTA model is validated in a standard floor high-precision ground construction task; results show that (1) the method can efficiently find potential optimal solutions based on the genetic algorithm; and (2) the strategic minimization of nonessential robotic deployment emerges as a critical factor in optimal task allocation. The technical innovation of the HRCTA model lies in systematically incorporating hazard levels into task allocation, enabling more adaptive and risk-aware decision-making in complex and evolving construction scenarios.