Many countries are making strategic plans to replace conventional vehicles (CVs) with electric vehicles (EVs), with the motivation to curb the growth of atmospheric CO₂ concentration. While previous publications have mainly employed social-economic based models to predict CO₂ emission trends from vehicles over the years, they do not account for the dynamics of engine and motor efficiency under different driving conditions. Therefore, this study utilized an experimentally validated vehicle dynamic model to simulate the consumption of gasoline and electricity for CVs and EVs, respectively, under eight driving cycles for different countries/regions. The CO₂ emissions of CVs and EVs through 2040 were then calculated, based on the assumptions of the improvement of engine efficiency and composition of power supply chain over the years. Results reveal that, assuming that the current projections and assumptions remain valid, China would have the highest CO₂ emission for EVs, followed by Japan, world average and the EU, mainly determined by the share of fossil fuels in the power grid. As for the influence of road conditions, the CO₂ emission of CVs was found to be always higher than that of EVs for all countries/regions over the years. The difference is around 10–20% under highway conditions, and as high as 50–60% in crowded urban driving conditions.

Accurate lane detection is an essential function of dynamic traffic perception. Though deep learning (DL) based methods have been widely applied to lane detection tasks, such models rarely achieve sufficient accuracy in low-light weather conditions. To improve the model accuracy in foggy conditions, a new approach was proposed based on monocular depth prediction and an atmospheric scattering model to generate fog artificially. We applied our method to the existing CULane dataset collected in clear weather and generated 107,451 labeled foggy lane images under three different fog densities. The original and generated datasets were then used to train state-of-the-art (SOTA) lane detection networks. The experiments demonstrate that the synthetic dataset can significantly increase the lane detection accuracy of DL-based models in both artificially generated foggy lane images and real foggy scenes. Specifically, the lane detection model performance (F1-measure) was increased from 11.09 to 70.41 under the heaviest foggy conditions. Additionally, this data augmentation method was further applied to another dataset, VIL-100, to test the adaptability of this approach. Similarly, it was found that even when the camera position or level of brightness was changed from one dataset to another, the foggy data augmentation approach is still valid to improve model performance under foggy conditions without degrading accuracy on other weather conditions. Finally, this approach also sheds light on practical applications for other complex scenes such as nighttime and rainy days.

Autonomous exploration in unknown environment has remained challenging due to unexpected collisions, stuckness and slowdowns around obstacles. This paper reports a novel approach based on Signed Distance Field (SDF), to optimize path planning algorithms and autonomous exploration strategy for safe and adaptive navigation in search and rescue missions. A quantitative criterion is established for evaluating the safety of planned trajectories. Simulation results show that the proposed SDF-A* path planner outperforms traditional methods with a 30.10% increase in path safety (i.e. average distance from robot to obstacles) and a 64.11% reduction in time consumption; The proposed SDF-based Safe Autonomous Exploration Strategy, combined with SDF-A* path planner, outperform traditional methods, leading to significant increases (47.06%) in path safety and reductions (44.75% and 15.32%) in exploration time and path length, respectively. The viability, efficiency, and safety of the proposed methods are further validated through real-world experiments on a three-wheeled differential steering robot equipped with Jetson Nano and RPLIDAR-A3 lidar. Results show that the proposed approach adapts to different indoor environments and map configurations without prior parameter settings.