roadrunner-scenario-simulating

Call createSimulation(rrApp) first, then simulateScenario(rrApp, IsSteppingStart=true, IsBlocking=false), and advance with set(rrSim, SimulationCommand="Step"). Add pause(stepSize) after each Step because the command is asynchronous and commands are dropped without waiting.

Get actors with get(rrSim, "ActorSimulation"), which returns a cell array, then call getAttribute(actorSim, "Pose") or getAttribute(actorSim, "Velocity"). Dot-property access like actorSim.Pose does not exist, and index 1 is always the world actor.

drivingScenario belongs to Automated Driving Toolbox and is a MATLAB-native scenario tool with unrelated APIs. RoadRunner scenario simulation uses simulateScenario and createSimulation on the roadrunner object, and the two must never be mixed.

Yes, build a model with a root-level RoadRunner Scenario block plus Reader and Writer blocks, publish it with Simulink.publish.publishActorBehavior, and assign the behavior to the actor. Start co-simulation with set_param(model, SimulationCommand="start"), never sim().

The Step command is asynchronous, so issuing steps without waiting causes commands to pile up and be silently dropped. Add pause(stepSize) after each set(rrSim, SimulationCommand="Step") call.

Do not use it for launching RoadRunner, authoring scenarios, building scenes, or exporting trajectories to CSV. Observers are read-only and cannot call setAttribute, and actors are only queryable while the simulation is running or paused.