A real-time data-driven approach for enhancing micro-mobility safety

Micro-mobility is a transport mode that refers to small and low-speed personal vehicles including human- and electric-powered vehicles such as bicycles and e-scooters. Nowadays, this mode of transportation is gaining popularity but the safety of micro-mobility users remains a concern in urban environments. Therefore, these safety risks including poor infrastructure, rider behaviour, interacting with other road users and collisions limit the choice to travel by micro-mobility. This papers introduces a novel approach for enhancing micro-mobility and Vulnerable Road Users (VRUs) safety. This approach is based on collecting real-time micro-level data about micro-mobility and rider’s behaviour by using a survey and developed sensory kit. Effectively, the approach involves three data-driven safety components or services which use the collected data to generate a micro-mobility hazards map and safe speed limit map, and to evaluate the effectiveness of protective equipment. Initially, machine learning technologies are used to train the collected data by the sensory kit and generate several classifiers responsible for detecting micro-mobility hazards and safe speed limits. Additionally, the correlation between the different variables of survey data is calculated to assess the efficiency of protective equipment. As a result, the proposed approach supports micro-mobility users in their day-to-day travelling and enhances their safety by enabling them to avoid dangerous locations, advising them to regulate their speeds when approaching hazards, and broadening their awareness about the benefits of protective equipment.