Many people share e-scooter rides which makes them a very dangerous vehicle. Therefore, detecting the weight and the number of people on a scooter before the ride begins is extremely important.
WhatI analyzed the data coming from a set of sensors installed on the scooters. I built a time-series analysis pipeline and a model to learn passenger weights and whether two people were sharing a ride.
HowThe project is implemented in pure Python. SKlearn has been used for the model and TSFresh for time-series windowing. Sensor signal analysis has been performed with Numpy and Scipy.