Demand forecasting for Hopp's e-scooter fleet

• Main Authors: Andri Fannar Pétursson 1995-, Egill Smári Snorrason 1998-, Halldór Holgersson 1995-, Jóhann Ingi Skúlason 1996-, Jón Skeggi Helgason 1999-
• Other Authors: Háskólinn í Reykjavík
• Format: Thesis
• Language: English
• Published: 2022
• Subjects: Tölvunarfræði; Vélrænt nám; Eftirspurn; Spálíkön; Rafhlaupahjól; Computer science; Machine learning; Supply and demand; Forecasting; Mathematical models; Scooters; Electric vehicles; Iceland
• Online Access: http://hdl.handle.net/1946/41884

Cities all around the world are dealing with the adverse effects of automobile travel and are therefore moving towards a more sustainable urban transportation system. Hopp is a micromobility platform launched by a small team of software developers in Iceland in 2019. In this project, we designed, implemented, and tested a machine learning tool that predicts the locational demand for Hopp's e-scooters in the near future. This tool aims to help reduce expenses, increase the number of rented scooters, and improve the reliability of the service.