Short-Term Demand and Volatility Prediction of Shared Micro-Mobility: a case study of e-scooter in Thammasat University

First-Mile/Last-Mile is one of the burdened urban transportation problems limiting the effectiveness of public transit. Recently, the newly emerged shared micro-mobility, escooter, came to fill this network gap and gained its popularity across the world. For this specific short-range trip mode, the demand is highly volatile from time to time, so demand and volatility prediction should be incorporated. For this reason, Box-Cox transformation, seasonal ARIMA (SARIMA), and family of GARCH models were used to predict its hourly demand and volatility, using hourly e-scooter demand from 23 Jan to 30 Apr 2019 in Thammasat University, Thailand. The combination of these models provides a very important understanding of the demand pattern that is necessary for operational planning.