In recent years transport of people and goods increased and traditional forms of transportation struggle to keep up with the demand. To keep cities' mobility systems healthy we investigate sustainable and safe alternatives.
Automated verhicles are one of the possible solutions, they can enable the redesign of urban built infrastructure and contribute to a sustainable and efficient mobility system. However, there are important factors to consider. For example, what would you do if a self-driving van approached at full speed? Do you cross over first or let the vehicle pass?
Goal and Research questions
SIPCATs goal is to find out under which conditions pedestrians and cyclists can safely share the road with automated verhicles, while both automated vehicles and other road users are going at an acceptable speed. The research questions that the project addresses are:
(1) what are the traffic safety risks in terms of hazard definition and characterization, contributing factors, and consequences?
(2) what are suitable objective surrogate measures of safety to represent these risks?
(3) what is the suitability of VR to investigate these risks in terms of relative validity and presence?
To investigate the interaction between autonomous transportation and humans, we launched a VR experiment as part of the SIPCAT project. In this virtual environment, you are at Marineterrein Amsterdam with autonomous vehicles driving around the area.
“As a program developer of smart urban mobility, I am very much involved in the question of the potential of a virtual world to make our traffic safer and more efficient.”
Tom Kuipers
Program Developer Smart Urban Mobility
Reducing barriers by using virtual reality
Real-world experiments can expose participating pedestrians and cyclists to dangerous situations and so they need to be accompanied by extensive, and sometimes, expensive safety measures. This is a barrier to the use of these experimental solutions. SIPCATs project team found a way around this issue by developing a 3D Virtual Reality (VR) Digital Twin.
This way researchers can safely and efficiently study the interaction of pedestrians and cyclists with automated shuttles and delivery robots. The team studies and develops use cases for pedestrians and cyclists interacting with the Husky delivery robot and the autonomous shuttle bus within the 3D VR Digital Twin of Marineterrein Amsterdam Living Lab (MALL).
“What makes this experiment unique is that we test with multiple people in one VR environment. This has not been done before due to the complexity.”
Sander Oudbier
Program Manager Smart Urban Mobility
A breakthrough in experimenting with VR
VR environments have already been used in previous research to study the interaction between pedestrians and cyclists and automated verhicles. However, those studies mainly focus on the 'subjective perception' of the safety of automated vehicles and pedestrian/cyclist interaction.
SIPCAT develops a new VR-based method that addresses subjective risk levels of automated vehicles and pedestrian/cyclist interaction. The research will focus on subjective measures such as Level of Trust, Presence, Perceived Behavioral Control, and Perceived Risk.
An important part of the project is the actual experiment with people in the 3D VR environment. Human participants test different scenarios in which they interact with the autonomous vehicle. The experiment is conducted with two participants in one and the same scenario.
They will both interact with each other as well as with the autonomous vehicle. The complexity of the multi-level interaction between the two participants and the autonomous vehicle has not been thoroughly researched and is a breakthrough in experimenting in virtual reality.
“Some traffic situations are too dangerous to test with real people. A digital twin offers a solution. With this experiment we test, for example, how people react to autonomous transport and what the interactions might look like. How does the autonomous van react when someone crosses the road? We can investigate this virtually without risk.”
Tom Kuipers
Program Developer Smart Urban Mobility
Eye contact with an autonomous bus
In addition to the aforementioned, a virtual environment enables us to easily test a great variety of situations. For example, is it safer for an autonomous vehicle to give a light signal to indicate that it has "seen" a pedestrian? Or can the vehicle project a crosswalk on the streets as an indication that it gives way to you?
When vehicles include a driver, as a pedestrian, you always check to see if this driver has seen you. With autonomous transport that is not possible. In the digital twin, we experiment with alternatives to this eye contact.
“The possibilities are endless, this experiment is only the beginning. Especially in the field of safety of autonomous transport, there is a lot of demand for information from government organizations. With this experiment we hope to answer a number of questions.”
Sander Oudbier
Program Manager Smart Urban Mobility
“This virtual world empowers us to model and simulate all kinds of situations, then involve real people and study how they react. We can apply that knowledge to the real world. We foresee that our future cities include autonomous transport of people and goods. We'd better be well prepared.”
Tom Kuipers
Program Developer Smart Urban Mobility
Results and lessons learned
In July 2021 a diverse consortium of partners set out to work on this project. This consortium consists of AMS Institute, TU Delft, TNO, Marinneterrein, Future Mobility Network, SWOV, Rijskdienst voor het Wegverkeer and VRBASE.STUDIO BV.
Results and lessons learned in SIPCAT will be used to develop other VR Digital Twin activities at MALL in the areas of energy, water, and urban planning as well as automated transport applications at other locations in the Netherlands.
