Exploring Spatiotemporal Data in Designing Human-Centric Perception for Autonomous Vehicles

About the Project

Autonomous vehicles (AVs) present several challenges for human-machine interaction that extend beyond the vehicle's interior to the surrounding street environment. While technical advancements in sensing, computer vision, and prediction have taken significant steps, these systems still struggle with the complexity of human behaviour. For example, Galvao et al. (2021) demonstrate that pedestrian and vehicle detection models remain limited in busy environments, while Mirzabagheri et al. (2025) highlight challenges in capturing subtle interactions, such as when pedestrians stop suddenly to interact or make sudden navigational changes. These gaps raise questions about how 'the human' is represented in AV perception and decision-making and how design and HCI perspectives might contribute to more responsible models of intent, navigation and interaction. Possible directions for this project include studying how different representations of human intent and navigation shape system design, prototyping forms of interactions between AVs and pedestrians, and developing frameworks for socially responsible data integration in AV perception.

Funding Notes

there is no funding for this project

References

1. Galvao, L.G., Abbod, M., Kalganova, T., Palade, V. and Huda, M.N., 2021. Pedestrian and vehicle detection in autonomous vehicle perception systems—A review. Sensors, 21(21), p.7267.
2. Mirzabagheri, A., Ahmadi, M., Zhang, N., Alirezaee, R., Mozaffari, S. and Alirezaee, S., 2025. Navigating Uncertainty: Advanced Techniques in Pedestrian Intention Prediction for Autonomous Vehicles—A Comprehensive Review. Vehicles, 7(2), p.57.