Among the key technologies of Industry 4.0, Augmented Reality (AR) is one of the most promising and enabling technologies for supporting factory workers and engineers at the workplace. To this end, the paper proposes a novel AR tool to assist operators during the inspection activities for the detection of production and assembly errors. In fact, thanks to the superimposition of the 3D models, as designed by the technical office, a worker can easily detect the presence of design discrepancies on the final physical assembled product and report them by adding 3D annotations directly on virtual models.

This AR tool has been developed by using ARCore™ libraries to ensure, in the first place, its compatibility with commonly used devices for which workers are already trained and, secondly, to take advantage of the hybrid-tracking techniques that combine vision- and sensor-based methods to improve the reliability of the AR visualization. Nevertheless, the proposed AR tool adopts multiple markers to minimize tracking errors and therefore to provide greater freedom of movement to the user, who can use the tool also for the assessment of large-size products.

Field experimentations have been carried out on a real case study with end-users in order to assess its usability and perceived mental workload through the SUS (System Usability Scale) and NASA-TLX (Task Load Index) standard questionnaires, respectively. The usability study was performed taking into account also objective metrics, i.e., by analysing user performance in target acquisition tasks while interacting with the AR tool. Statistical analysis proved that the adoption of this AR tool requires low mental demand, and its usability has reached a high level of satisfaction both by the factory workers and engineers involved in the user study.

在工业4.0的关键技术中，增强现实（AR）是支持工作场所的工厂工人和工程师的最有前途和使能的技术之一。为此，本文提出了一种新颖的AR工具，可在检查活动中协助操作员检测生产和装配错误。实际上，由于技术办公室设计了3D模型的叠加，工作人员可以轻松地检测最终物理组装产品上是否存在设计差异，并通过在虚拟模型上直接添加3D注释来报告这些差异。

通过使用ARCore™库开发此AR工具，首先要确保其与已经对其进行了培训的常用设备兼容，其次，要利用结合了视觉和传感器功能的混合跟踪技术基于方法来提高AR可视化的可靠性。尽管如此，建议的AR工具采用了多个标记，以最大程度地减少跟踪误差，从而为用户提供更大的移动自由度，用户也可以使用该工具来评估大型产品。

为了对最终用户分别通过SUS（系统可用性量表）和NASA-TLX（任务负载指数）标准调查表评估其可用性和感知的工作量，对最终用户进行了实地实验。进行可用性研究时还考虑了客观指标，即通过与AR工具交互时分析目标获取任务中的用户性能。统计分析证明，采用此AR工具所需的心理需求较低，并且其可用性已被参与用户研究的工厂工人和工程师所达到的高度满意。