Designing controlled experiments in extended reality typically requires proficiency in game engines like Unity or Unreal, which limits many transportation researchers. XARA aims to remove that barrier by letting researchers use natural language and hand gestures to build and modify XR scenarios.
The system's pipeline first transcribes spoken commands, then a dual large language model architecture interprets them into structured commands that manipulate a Unity-based XR scene in real time. It supports object creation, spatial manipulation, animation control, and scene persistence through a modular architecture.
The authors describe potential uses in behavioral travel experiments, safety studies, human factors research, and policy testing. By lowering the technical threshold, XARA could enable rapid prototyping of driving, pedestrian, and urban design scenarios.