SunChoice: A Novel Framework for Sunshine-Based Trajectories Analysis

Trajectory analysis based on sunshine information shows that sunshine has a crucial effect on many conditions involving driving comfort and route planning. However, based on our best understanding, a little, if any, work takes into consideration the influence of sunshine on large numbers of trajectories of mobile objects such as vehicles. In this paper, we analyze, for the first time, the correlation between sunshine and large numbers of trajectories of mobile objects while presenting compression of the sunshine-based trajectories on a national scale. Within the research, a sunshine-oriented model is introduced to combine sunshine with real trajectories through the extraction of information centering around sunshine on arbitrary trajectories. Our trajectory analysis based on sunshine is versatile, and the model mentioned above is implemented to enhance the riding comfort of passengers. Our model is evaluated with an extensive simulation using general transit feed specification data over three months from August to October 2017 in five U.S. cities. We also conduct a field study in Minneapolis to show the validity and efficiency of our model. The evaluation results show that our design: 1) decreases trajectory computation by 88.7% with the trajectory compression and 2) achieves more than 90% accuracy for the trip sunshine information prediction in both the simulation and real-world tests.