Titulo Estágio
Development of a Virtual Reality Driving Simulator for Evaluating In-Vehicle Interfaces in Connected Vehicle Scenarios
Áreas de especialidade
Sistemas de Informação
Engenharia de Software
Local do Estágio
DEI/VRLab
Enquadramento
The adoption of intelligent and connected vehicles (ICVs) is reshaping transportation systems, with Vehicle-to-Everything (V2X) communication enabling new forms of situational awareness. One of the main challenges in this area is designing Human-Machine Interfaces (HMIs) that effectively inform the driver about critical events—such as overtaking situations or pedestrian crossings—without introducing cognitive overload or distraction.
To properly test and evaluate such HMIs under safe yet realistic conditions, virtual environments are increasingly used. Virtual Reality (VR) driving simulators offer a powerful platform for creating immersive and repeatable driving scenarios that include complex interactions, such as other vehicles transmitting sensor data or infrastructure alerts. However, general-purpose driving simulators often lack the flexibility to simulate the types of V2V, V2I, or V2X communication envisioned in real-world vehicular networks.
This thesis focuses on developing or adapting a VR driving simulator to support the simulation of connected vehicle scenarios. The simulator will serve as an experimental testbed for evaluating HMIs designed in parallel work. It must include components such as simulated communication with nearby vehicles, infrastructure elements (e.g., crosswalk sensors), and the capacity to trigger interface elements in response to real-time contextual changes.
The simulator will be used to replicate safety-critical use cases identified by the ADSafeVANET project, such as overtaking maneuvers and pedestrian crossings in low-visibility conditions. The goal is to support user studies that examine how drivers perceive and react to interface notifications in realistic, controlled simulations.
Objetivo
The primary goal of this thesis is to design and implement a flexible, extensible VR-based driving simulator tailored for evaluating in-vehicle HMIs in connected vehicle environments.
Specific objectives include:
- Investigate existing VR and non-VR driving simulators (e.g., open-source options like CARLA, OpenDS, or Unity-based systems), focusing on adaptability and support for V2X simulation.
- Translate the two safety-critical ADSafeVANET use cases—overtaking and pedestrian crossings—into simulation requirements including environmental conditions, vehicle behavior, and communication triggers.
- Develop driving scenarios with controllable vehicle traffic, weather/visibility conditions, and pedestrian behavior. The environment must support a VR headset and input/output integration for interface testing.
- Implement simplified V2V and V2I mechanisms (e.g., simulated messages about blind spot vehicles or pedestrian detection) that dynamically affect the environment or trigger driver notifications.
- Provide API hooks or modules to integrate and test different HMIs (visual/auditory/haptic), either developed within this thesis or imported from parallel design work.
- Implement data logging (e.g., events, reactions, timings) to support usability and UX evaluation during user studies.
- Conduct a small-scale pilot test to evaluate the realism, responsiveness, and usability of the simulator for both participants and researchers.
Plano de Trabalhos - Semestre 1
- Conduct a literature review on existing VR driving simulators and simulation frameworks for vehicular networks (e.g., Unity, CARLA, OpenDS).
- Explore case studies where simulators have been used to evaluate HMIs.
- Define technical and functional requirements based on the two main use cases (overtaking, pedestrian crossing).
- Select the base simulation platform (likely Unity for flexibility and integration with VR headsets).
- Begin initial prototyping of the driving environment (urban road with crosswalks and overtaking zones).
- Research methods for simulating communication events (e.g., vehicles broadcasting positions, infrastructure detecting pedestrians).
- Implement basic driving mechanics and environmental elements (e.g., road, vehicles, crosswalks, foggy conditions).
- Define the simulation architecture, including modular components for adding HMIs and communication events.
- Develop a preliminary proof-of-concept (PoC) driving experience using VR (e.g., Oculus Quest or HTC Vive).
- Document system design and development choices.
- Begin writing initial chapters of the thesis (Introduction, Related Work, System Architecture).
Deliverables for First Semester:
- Literature and technology review
- Initial simulator prototype with one sample scenario
- Design documentation of the simulation architecture
- Interim thesis
Plano de Trabalhos - Semestre 2
- Add support for simulated V2V and V2I communication events:
* E.g., simulate a vehicle ahead sending blind spot alerts
* E.g., simulate a roadside unit broadcasting a pedestrian crossing event
- Implement API or event hooks to allow HMI modules to subscribe to these events.
- Integrate one or more example HMIs into the simulator.
- Implement logging of key metrics (reaction time, gaze direction, event triggers).
- Prepare and conduct pilot user test with 3–5 participants to assess simulator’s usability and usefulness.
- Refine the simulator based on feedback.
- Finalize the implementation of all planned features.
- Write results and discussion chapters of the thesis.
-
Deliverables for Second Semester:
- Fully working VR driving simulator for connected vehicle scenarios
- Logging and integration features for interface testing
- Final thesis
Condições
The student will have access to all the computational resources necessary to carry out the work.
A workspace at CISUC (Centre for Informatics and Systems of the University of Coimbra) will also be made available.
There is the possibility of funding in the ADSafeVANET project (task T2).
Orientador
Jorge C. S. Cardoso
jorgecardoso@dei.uc.pt 📩