Mobility Choices is used to predict transportation mode choices of students at the University of Guadalajara (UdeG) within the Guadalajara Metropolitan Area in Mexico. In this case, there will be a need to address limited data, which is uncommon for this type of study, and the prevalence of informal transportation, common in regions of Latin America such as Guadalajara.
Our daily mobility choices can be influenced by a considerable number of factors: age, weather, travel time, city, income, fastest mode, parking availability, etc. City Science has developed a tool to predict our mobility choices by incorporating all these factors into a Machine Learning model. In this case, thanks to our City Science Network, we have been able to apply this tool in the Metropolitan Area of Guadalajara through a student survey. However, applying this tool in such territories in Latin America is challenging due to the prevalence of informal transportation in the area. Furthermore, the lack of telecom data, as seen in other case studies, complicates the study.
Due to the large size of the UdeG, we have selected as a case study all individuals frequenting student centers in Guadalajara and Zapopan, the two largest municipalities in the Metropolitan Area. This approach will aid in predicting movements around the University Cultural Center in Zapopan,situated near the border with the city of Guadalajara. This center is currently under development and is anticipated to experience a significant increase in population by 2050.
In today's society, student transportation constitutes the second-largest share of peak hour traffic. In our study, approximately 75% of students predominantly use public transportation, with private autos following at around 15%. Walking and cycling represent approximately 9% and 2%, respectively.
Our machine learning model can predict the Mobility Choice with an accuracy of 75%. Additionally, we are able to forecast potencial scenarios (What-if Scenarios) to help us in future decision-making. For the case study, students from the University Cultural Center (UCC) have been taken into account.
It has been demonstrated that the closer we bring students to their student centers, the more we increase the population walking and the less tendency they have to use private cars. Furthermore, this tool's potential impact in the future is significant, given the substantial reduction in CO2 emissions it could bring about. By proposing potential scenarios that generate a positive impact, it actively promotes a shift towards more sustainable living. Additionally, there are a few conclusions to mention regarding this project:
Simulations and visualizations were conducted using Kepler, QGIS, and Jupyter Notebooks