A Data Science Framework to Measure Vehicle Miles Traveled by Mode and Purpose

Research Members: Albert Cao, Cristobal Pais

The outbreak of COVID-19 drastically changed people’s mobility patterns and travel behavior in California. Identifying the underlying patterns of these changes within the context of COVID-19 provides insights on designing policies that can help the State reach its 2030 greenhouse emission goals. With the power of anonymized Location-Based Service (LBS) data collected from mobile devices, (1) we found vehicle usage decreased up to 20% more in urban areas than in rural areas because of COVID-19. (2) The number of work commutes decreased 30% more than the number of non-commute trips. The number of commuting travels we observed in 2022 had not returned to the respective 2019 level, suggesting that remote work had a lasting impact in travel behavior. (3) In the two-week period following the beginning of the lockdown in March 2020, many more people changed residential locations compared to all other periods of observation in 2020. Additionally, we developed an unsupervised mode detection scheme through which we found that the introduction of electric scooters in 2019 in Sacramento decreased the vehicle usage rate.


Marta C. González, Shangqing Cao, Cristobal Pais, Violet Lingenfelter, Ruining Wang, Elouan Pulveric, and Tomas Wenzel. “A Data Science Framework to Measure Vehicle Miles Traveled by Mode and Purpose.” Report to California Air Resources Board Research Division (2024). [Project Page] [PDF]

HuMNet Lab