In order to provide safe development of road freight traffic, this paper proposes a truck driving risk identification method based on Optuna optimization of machine learning model. First, the risk characterization indicators were extracted from the natural driving data of trucks, and the threshold value of each indicator was determined using a box plot-based method. Second, the truck driving risk was quantified into three categories of low level, medium level, and high level risk, and the unbalanced data were processed using a hybrid sampling algorithm. Finally, the tree-based decision tree (DT) model, random forest (RF) model, Light Gradient Boosting Machine (LightGBM) model and eXtreme Gradient Boosting (XGBoost) model were selected for training and Optuna was used for hyperparameter optimization of the model. The results are shown to indicate that the machine learning model based on Optuna optimization can effectively identify truck driving risks. Combining the running time, precision, recall, and F1-Score, the LightGBM model optimized based on the Tree-structured Parzen Estimator (TPE) algorithm has the best performance with a precision of 0.98. In addition, the speed mean has the highest feature importance of 14%, which needs to be focused on when preventing truck driving risks. The research results can provide policy support for transportation management departments to formulate risk control measures for trucks.

Purpose: COVID-19 has aggregated the need for a non-contact medical logistic system. A non-contact robot with self-navigation ability has greatly enhanced the efficiency of the medical logistic system. This paper aims to design a new medical logistics robot system for the complex environment of hospitals with dynamic obstacles. Design/methodology/approach: Targeting a medical logistics robot system for a large-scale hospital environment, this study proposed a dynamic obstacle avoidance system to reduce the robot’s delay time as well as frequent route switching. In the algorithm, this study proposed a new loop closure detection with an artificial correction factor. Moreover, this study presented enhanced 3D object detection, improving detection accuracy in hospital environments. Findings: Experimental results confirm that the robot can move along its global path and reach its destination without colliding with stationary or moving obstacles. Originality/value: The medical logistics robot system has safe and stable performance in real hospital scenarios. The implementation verifies that the robot has effectiveness and reliability in both hardware and software design.

Bikeshare, as a convenient transport mode, can address the first- and last-mile travel needs of metro trips while generating many environmental benefits, such as reducing the use of environmentally unfriendly transport modes and lowering the carbon emissions of the urban transportation system. This paper takes bikeshare as a feeder mode of metro stations (BS-FMMS) as the research object and compares the spatial and temporal differences in the carbon emission reduction benefits of BS-FMMS on workdays and non-workdays by using the framework of BS-FMMS carbon reduction benefit analysis and the methods of time-series analysis, spatial aggregation analysis, and box plot analysis. The results show that the carbon emission reduction benefit of bikeshare has obvious morning and evening peaks on workdays, while it tends to be stable without obvious peaks during the day on non-workdays. From the perspective of spatial distribution, the carbon emission reduction benefits of BS-FMMS are more significant in the metro station areas in the south of Baoan district, the west of Nanshan district, the central of Longhua district, and the south of Futian district in Shenzhen city, and the metro stations where the carbon emission reduction benefits of the non-workday are greater than those of the workday are mainly concentrated in Nanshan district, Futian district, and Luohu district. There is a significant positive correlation between BS-FMMS ridership and carbon emission reduction. These findings can provide clear policy implications for the decarbonization development of urban transportation systems.

Clarifying the spatiotemporal evolution and driving mechanisms of bike-sharing-to-metro feeder trips (BSMF) is key to optimizing urban public transport’s first-and-last-mile connectivity and advancing low-carbon development. Existing studies on BSMF mostly ignore spatiotemporal heterogeneity, lack in-depth exploration of multi-factor interaction effects, and have subjective stratification or model specification bias, which hinder the accurate depiction of BSMF’s complex evolutionary patterns. Taking Xi’an as a case with 126 metro stations as analysis units, this study integrates multi-source data including shared bike trip records, metro network and built environment attributes to address the above issues. A framework combining kernel density estimation, spatial autocorrelation analysis, Optimal Parameter Geographic Detector (OPGD) and Geographically and Temporally Weighted Regression (GTWR) models (OPGD-GTWR) is constructed to identify BSMF’s spatiotemporal patterns, screen key influencing factors and reveal their spatiotemporal heterogeneity and interactive mechanisms. Results show Xi’an’s BSMF trips feature a “double-peak and double-valley” temporal tidal pattern and core-periphery spatial agglomeration. The OPGD-GTWR model (R² = 0.853) outperforms traditional models in capturing spatiotemporal heterogeneity. These findings provide empirical evidence and refined references for shared mobility resource allocation, bike-metro integration improvement and transit-oriented urban planning.