VirtuAV – A safe virtual environment to test critical interactions between vulnerable road users and automated vehicles with high ecological validity

Forskningsprojekt, 2027

Safe interaction between automated vehicles (AVs) and vulnerable road users (VRUs) is essential for
deploying vehicle automation in urban environments. Real‑world tests of AV–VRU encounters are
costly and constrained by ethical and safety considerations. Virtual environments (VEs) offer safety,
control, and repeatability, even for safety‑critical scenarios. Typical quality measures of VE are
presence, immersion, and awareness. Each measure addresses critical aspects of the virtual
experience, including the real-life relevance of the experience (that is, ecological validity).
High‑fidelity VEs such as the HIKER simulator enable VRU studies by allowing participants to move
freely within a large physical space while immersed in a projected virtual environment. Although
technologically advanced, such systems are expensive, limited to pedestrian users, and not common
in Europe. At Chalmers, existing motion‑capture infrastructure and in‑house expertise may allow us
not only to replicate HIKER‑like functionality, but also to extend it to micromobility users, including
cyclists and e-scooterists, creating a unique asset for future (European) research projects.
The proposed concept is straightforward: track a pedestrian or rider using Chalmers’ motion‑capture
system in Eventhallen and render their position in real time via a VR headset. For example, a cyclist
riding within Eventhallen could experience controlled, safety‑critical encounters with AVs in a virtual
intersection while maintaining natural kinematics and balance, thereby preserving ecological validity.
However, several non‑trivial engineering and human‑factors challenges may hinder feasibility. These
include: (1) the potential computational and communication burden of real‑time virtual reality at full
scale, (2) visual feedback delays that may compromise rider balance, (3) susceptibility to simulator
sickness, (4) stringent robustness requirements to ensure participant safety in the presence of even
minor technical failures, and (5) uncertainty regarding whether ecological validity will be sufficient
for meaningful real‑world inference. This project seeks to commence the development of this VRbased
test facility at Chalmers by investigating the issues above and preparing a plan (including
projects applications) to make this test facility a reality while also involving new stakeholders and
UCs.