The good news is that this matches my intuition from the visualization. However, some weakness is to be considered. First, the distribution of trip duration, as seen in Fig 1, is more identical to a Poisson or chi-square distribution. I may need to revisit the assumption of the distribution being Gaussian. The second improvement to be made is to include data from different months. Using data from winter, in which people use bikes less, may fail to capture the trip pattern in other seasons.