Portobello: Extending Driving Simulation from the Lab to the Road

TL;DR

Portobello addresses the challenge of porting driving-study designs between in-lab and on-road XR platforms to improve ecological validity. It introduces a LiDAR-based, map-centric localization stack and a common modeling framework to twin studies across platforms, demonstrated by a within-subjects crosswalk-cooperation study ($N=32$) run on both in-lab and on-road setups. The contributions include the Portobello infrastructure, integration with XR-OOM, and methodological guidelines for study design portability; results show mixed but informative platform effects, e.g., on-road increases natural head movements and perceived task realism, while in-lab offers lower discomfort and higher perceived immersion. The work advances open-science by releasing Portobello code and providing a blueprint for twinned studies to enhance ecological validity and translational impact in automotive HCI research.

Abstract

In automotive user interface design, testing often starts with lab-based driving simulators and migrates toward on-road studies to mitigate risks. Mixed reality (XR) helps translate virtual study designs to the real road to increase ecological validity. However, researchers rarely run the same study in both in-lab and on-road simulators due to the challenges of replicating studies in both physical and virtual worlds. To provide a common infrastructure to port in-lab study designs on-road, we built a platform-portable infrastructure, Portobello, to enable us to run twinned physical-virtual studies. As a proof-of-concept, we extended the on-road simulator XR-OOM with Portobello. We ran a within-subjects, autonomous-vehicle crosswalk cooperation study (N=32) both in-lab and on-road to investigate study design portability and platform-driven influences on study outcomes. To our knowledge, this is the first system that enables the twinning of studies originally designed for in-lab simulators to be carried out in an on-road platform.

Figures (5)

• Figure 1: Complete execution pipeline using the Portobello infrastructure. During the design phase, researchers can drag and drop virtual objects on the point cloud map. At run time, the LiDAR-based navigation system locates the vehicle's position within the same map and sends the relative vehicle position to Unity. As a result, passengers wearing the video-see-through headset can see virtual objects at their corresponding real-world locations.
• Figure 2: The Portobello system uses a LiDAR-based navigation stack to localize the vehicle's runtime position within a given map. The location information is transmitted to the Unity Desktop through ROS. The black dotted line indicates a virtual divider appearing only in the headset.
• Figure 3: Our study area for on-road simulator on Roosevelt Island . The pre-determined test route is highlighted in blue. The crosswalks with staged interactions are highlighted in the red bounding boxes. The start and end locations are denoted by the green and purple dots, respectively.
• Figure 4: Participants' ratings of each simulator on a five-point Likert scale across seven distinct dimensions where 1=low and 5=high.
• Figure 5: After experiencing both simulators, participants also directly compared the in-lab simulator and the on-road simulator on the same Likert scale. Participants' responses are shown as histograms.