D2.4 Definition of the future use cases: scope and data to build digital twins of use cases

The SAFE-UP project aims to proactively address the novel safety challenges of the future mobility systems through the development of tools and innovative safety methods that lead to improvements in road transport safety. Future mobility systems will rely on partially and fully automated vehicles to reduce traffic collisions and casualties by removing causal factors like driver distraction, fatigue or infractions and by reacting autonomously to emergency situations. On the other hand, they may introduce new collision risk factors or risky behaviours in other traffic participants. SAFE-UP’s Work Package 2 will further the understanding of the impact of vehicle automation technologies on safety by leveraging newly developed behavioural traffic simulation tools. These tools will allow one to simulate specific road networks with a variable proportion of automated vehicles. The simulation and prediction of future safety-critical scenarios requires development of a new traffic simulation environment and framework which deals with the specific challenges around road collisions. From the Grant agreement: “D2.4: Definition of the future use cases: scope and data to build digital twins of use cases”, this Deliverable presents the SAFE-UP traffic simulation environment with the next generation of road users’ behavioural models in the road network before and after the introduction of autonomous vehicles (AV)s. The first model simulates the human-driven vehicle with two-dimensional manoeuvres and in-lane interactions with cyclists; the second model simulates automated driving behaviour with 2D-trajectory planning controlling longitudinal and lateral movements, and the third group of models simulates he behavioural models for VRUs such as cyclists, pedestrians, and powered two wheelers (PTWs). This new generation of driving behavioural models together with new safety metrics will be systematically integrated in the Aimsun Next traffic simulation platform. The integration framework is presented in this deliverable. This framework enables harmonised simulation of the next generation of all road users’ behavioural models (driver, AV, and VRU), capturing the failure of sensors, the errors of judgement that drivers and riders might take, and their distracted perception in the traffic conditions we know today and future with the presence of AVs. Furthermore, the deliverable presents the methodology and data requirements for the calibration and validation of the behavioural road users’ models and traffic network model in three network scenarios. The methodology is designed to work with various data sources for the calibration and validation of the models and traffic network, to ensure a reliable simulation output. In this deliverable, Section 2 presents an approach adopted to build the road network representation (static objects) and traffic conditions (dynamic or moving objects) that the virtual scene (digital twin) for simulation of all road users and AVs. Section 3 presents a summary of the behavioural models for all road users developed in Task 2.3 and integration framework used to create co-simulation environment of all agents. Section 4 gives a detailed integration framework description and covers all required integration features. Section 5 provides data requirements for calibration and validation of the key behavioural model parameters and the network model, to ensure a close representation of traffic conditions and road users interactions, including safety critical interactions. Finally, Section 6 provides our conclusions and recommendations.