The application of augmented reality-based building defect inspection (AR-BDI) has become increasingly common in construction practice. Visual fatigue and related discomfort remain practical ergonomic concerns during prolonged AR head-mounted display (AR-HMD) use, which may negatively affect worker performance and safety. To address this ergonomic challenge, this study proposes an integrated method that combines human–computer interaction (HCI) interface optimization with work–rest scheduling based on eye tracking data. A total objective function, developed from visual attention mechanisms, is optimized through a genetic algorithm (GA) with a penalty mechanism to improve interface layout efficiency. In addition, an LSTM-based model is used to monitor visual fatigue levels in real-time and dynamically adjust rest intervals. Field experiments with AR-HMDs indicate that the optimized HCI interface enhances information recognition efficiency and reduces eye movement load. The proposed scheduling system further supports visual fatigue management by enabling timely rest interventions. By integrating ergonomic interface design with sequential visual fatigue prediction and fatigue-aware work–rest scheduling in AR-BDI tasks, this study offers a structured method to enhance user experience and long-term system usability in construction settings.

Traffic congestion has become a worldwide problem, seriously restricting the performance of transport network. Association of congested links is an essential factor in the formation of traffic congestion, while it lacks enough research. In this paper, a methodology ARM-K combining K-means clustering and association rules mining is proposed to discover the relations of congested links. Specifically, the method digs out the association of two congested links and further extends the 2-tuple relations into a relation graph. As a result, the relation graph and its topological-order profile the associations of congested links. The experiments including congestion prediction and congestion dispersion are devised for the proposed method based on mobile data derived from the traffic simulator VISSIM. The results show relatively high prediction accuracy (above 0.75) and obvious improvement of network performance after dispersion (more than 12.5 percentage-point drop of travel time index), indicating that ARM-K can effectively explores the associations of congested links and provide valuable information for congestion management.