Augmented Reality Head-Up Displays (AR-HUDs) are increasingly studied as next-generation in-vehicle interfaces, yet findings remain difficult to reproduce and compare because evaluations rely on heterogeneous simulation environments and experimental protocols. A shared, openly reusable immersive infrastructure for controlled AR-HUD experimentation is still lacking. We present CARLA-IVR, an open extension of the CARLA simulator that adds stereoscopic XR rendering, embodied interaction with the vehicle, and a scene-referenced AR-HUD overlay layer while preserving interoperability with standard CARLA workflows. We technically characterize the closed-loop pipeline through a rendering throughput profile, an input-to-render latency budget, and a scene-population ramp that delimits the real-time operating envelope of our platform, without human participants. We then present the platform through an initial, literature-grounded experimental design comparing AR-HUD and head-down display (HDD) conditions across two simulator setups: an immersive configuration (IVR) and a desktop configuration (DVR). Objective and subjective measures were collected from 34 participants. AR-HUD-HDD differences were descriptively more pronounced under IVR for perceived workload and trajectory excursions, whereas DVR showed weaker differences. However, the available sample size did not provide statistical support for a reliable interaction between interface type (AR-HUD vs. HDD) and simulator setup (IVR vs. DVR), and the IVR and DVR setups differed in several respects beyond visual immersion. The user study is therefore interpreted as exploratory evidence and a reproducible evaluation case rather than as confirmatory evidence of AR-HUD efficacy. Regardless, CARLA-IVR provides an open, interoperable simulation foundation that supports reproducible AR-HUD evaluation and the development of more systematic, comparable immersive AR-HUD evaluation protocols.

CARLA-IVR: An Open CARLA Extension for Immersive AR-HUD Evaluation and Reproducible Experimental Design / Vezzani, L., Strada, F., Bottino, A.. - In: IEEE OPEN JOURNAL OF VEHICULAR TECHNOLOGY. - ISSN 2644-1330. - ELETTRONICO. - (2026), pp. 1-16. [10.1109/ojvt.2026.3740837]

CARLA-IVR: An Open CARLA Extension for Immersive AR-HUD Evaluation and Reproducible Experimental Design

Vezzani, Leonardo;Strada, Francesco;Bottino, Andrea
2026

Abstract

Augmented Reality Head-Up Displays (AR-HUDs) are increasingly studied as next-generation in-vehicle interfaces, yet findings remain difficult to reproduce and compare because evaluations rely on heterogeneous simulation environments and experimental protocols. A shared, openly reusable immersive infrastructure for controlled AR-HUD experimentation is still lacking. We present CARLA-IVR, an open extension of the CARLA simulator that adds stereoscopic XR rendering, embodied interaction with the vehicle, and a scene-referenced AR-HUD overlay layer while preserving interoperability with standard CARLA workflows. We technically characterize the closed-loop pipeline through a rendering throughput profile, an input-to-render latency budget, and a scene-population ramp that delimits the real-time operating envelope of our platform, without human participants. We then present the platform through an initial, literature-grounded experimental design comparing AR-HUD and head-down display (HDD) conditions across two simulator setups: an immersive configuration (IVR) and a desktop configuration (DVR). Objective and subjective measures were collected from 34 participants. AR-HUD-HDD differences were descriptively more pronounced under IVR for perceived workload and trajectory excursions, whereas DVR showed weaker differences. However, the available sample size did not provide statistical support for a reliable interaction between interface type (AR-HUD vs. HDD) and simulator setup (IVR vs. DVR), and the IVR and DVR setups differed in several respects beyond visual immersion. The user study is therefore interpreted as exploratory evidence and a reproducible evaluation case rather than as confirmatory evidence of AR-HUD efficacy. Regardless, CARLA-IVR provides an open, interoperable simulation foundation that supports reproducible AR-HUD evaluation and the development of more systematic, comparable immersive AR-HUD evaluation protocols.
File in questo prodotto:
File Dimensione Formato  
CARLA-IVR_An_Open_CARLA_Extension_for_Immersive_AR-HUD_Evaluation_and_Reproducible_Experimental_Design.pdf

accesso aperto

Tipologia: 2. Post-print / Author's Accepted Manuscript
Licenza: Creative commons
Dimensione 4.56 MB
Formato Adobe PDF
4.56 MB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/3016913