Department of Architectural Engineering, The Pennsylvania State University, University Park, PA, 16802, USA.
Department of Mathematics, Chariho Regional High School, Richmond, RI, 02894, USA.
Sci Rep. 2023 Feb 20;13(1):2929. doi: 10.1038/s41598-023-29950-w.
This work seeks to capture how an expert interacts with a structure during a facade inspection so that more detailed and situationally-aware inspections can be done with autonomous robots in the future. Eye tracking maps where an inspector is looking during a structural inspection, and it recognizes implicit human attention. Experiments were performed on a facade during a damage assessment to analyze key, visually-based features that are important for understanding human-infrastructure interaction during the process. For data collection and analysis, experiments were conducted to assess an inspector's behavioral changes while assessing a real structure. These eye tracking features provided the basis for the inspector's intent prediction and were used to understand how humans interact with the structure during the inspection processes. This method will facilitate information-sharing and decision-making during the inspection processes for collaborative human-robot teams; thus, it will enable unmanned aerial vehicle (UAV) for future building inspection through artificial intelligence support.
这项工作旨在捕捉专家在进行立面检查时与结构的交互方式，以便未来能够使用自主机器人进行更详细和情境感知的检查。眼动追踪地图记录了在结构检查期间检查员的视线位置，并识别出隐含的人类注意力。在进行损伤评估时，对面部进行了实验，以分析在该过程中理解人类与基础设施相互作用时很重要的关键基于视觉的特征。为了进行数据收集和分析，进行了实验以评估检查员在评估真实结构时的行为变化。这些眼动追踪特征为检查员的意图预测提供了基础，并用于了解人类在检查过程中如何与结构相互作用。这种方法将促进在检查过程中的信息共享和决策制定，以实现人机协作团队；因此，它将通过人工智能支持使未来的建筑检查能够使用无人机 (UAV)。
