Interactions Between Automated Shuttle and Vulnerable Road Users: A Case Study

The evermore introduction of automated systems in driving operations opens the possibility of an upcoming traffic scenario with automated vehicles. Research about automated vehicles in traffic has been run with automated shuttles mainly for two reasons: to provide a wide idea about users’ acceptance (more users can get on a bus than on a private vehicle) and valuable simulation of interactions with other vehicles, but on a fixed and limited route. The main concern in this sense is a quantitative estimation (speed, acceleration, and deceleration) of the automated shuttle while interacting with vulnerable road users (VRUs) and regular vehicles. This specific analysis about the interactions was the main focus of this study. The proposed study was a real-world simulation set in Bari (Italy), with a partially automated shuttle bus provided by Navya. Data about the vehicle interacting with pedestrians, e-scooters, and regular vehicles were collected and then processed. The automated vehicle behaved in a safe way in all the tested interactions, keeping its speed very low (below 15 km/h), to ensure comfort also during the emergency braking (with a maximum deceleration of 1.5 m/s²). The most outstanding result of the tests was to verify that the automated shuttle behaves in the same way for all the different interactions, regardless of the type of user interacting and the modality (crossing, overtaking, following). This result means a lot because it suggests that the vehicle can safely deal also with unpredictable decisions by VRUs, that otherwise could have constituted a big issue for the implementation of such types of vehicles.