Task-Driven Framework for Driving Simulation: Scenario Orchestration with Autonomous Simulated Vehicles

Scenarios in driving simulators cover what the human participants experience and what the researchers need: the physical scene, predefined traffic flow, simulated vehicles' interactions with the participants and measurements to be collected. Previous methodologies for orchestrating scenarios regarding the interactions have the following drawbacks: 1) action sequences that simulated vehicles should follow lack the contexts of each action; 2) programming languages always include platform-dependent details and are not suitable for context modelling and 3) scenarios cannot be generated dynamically to cope with failures that happen in trials. To overcome the limitations above, an Ontology for Scenario Orchestration (OSO) was first developed to model concepts and their relationships in the domain of scenario orchestration, including a concept named Assignment, which represents the task(s) of virtual drivers and encodes the contextual information of proposed actions, e.g., simulated vehicles involved. It can also provide a file for machine processing. An algorithm named NAUSEA (autoNomous locAl manoeUvre and Scenario orchEstration based on automated action plAnning) was generated to utilise Assignments recorded using OSO. Encoded in the driver model SAIL (Scenario-Aware drIver modeL), NAUSEA can be used by a virtual driver to control simulated vehicles dynamically. Failed Assignments, designed to generate specific interactions, can be re-tried if permitted. A framework SOAV (Scenario Orchestration with Autonomous simulated Vehicles) was formed to support SAIL/NAUSEA and orchestrate scenarios with autonomous vehicles. Three verification experiments showed that SOAV worked properly by generating desired interactions and dealing with failures. OSO can also provide contextual information in a human-readable and machine processable manner