Virtual Immersive Reality Environment (VIRE) for Disruptive Vehicular Technology Choice Experiments

Introduction In this study we developed a Virtual Immersive Reality Environment (VIRE) using Head Mounted Display (HMD) and employed it to analyse the choice behaviour of individuals in a future situation where the vehicles would be fully automated and connected to the network. In such situation, based on the changing multimodal traffic conditions, the transportation network can be controlled dynamically via vehicle-to-vehicle (v2v), vehicle-to-infrastructure (v2i), and infrastructure-to-infrastructure (i2i) communication. Thus, a system optimal would be most attainable, practical and will have various societal benefits e.g. lower congestion, efficient use of the system, and lower cost. Such strategy may have negative consequences on the individual users e.g. some users may experience longer travel times or they may not be taken on their preferred route. In particular, we are interested in systematically analysing the decision to select driverless car by the respondents on a multimodal network that is implementing a system optimal strategy using v2v, v2i, and i2i communications. Background The Discrete Choice Experiments (DCEs) research community proactively explores new tools in order to make the choice tasks and thus responses more realistic. Darbani et al. (2013) conducted location choice experiments using text and gaming engine based animations. They concluded that the combination of visual and textual information increased the processing of information by the respondents. Cherchi and Hensher (2015) pointed out the need for visual and engaging tools e.g. eyes tracking, virtual reality, and simulators to add value to behavioural relevance and reduce hypothetical bias. However, due to the amount of work needed to actually develop these tools, very little progress has been made in terms of the main stream usage of more immersive tools in DCE. This fact is evident from the recent literature on the investigation of the effects of automated vehicle technology. For instance, Bansal et al. (2016) and Daziano et al. (2016) used web-based survey tools to determine the willingness-to-pay of population in the U.S. for automated vehicles. An important dimension of disruptive technology like automated vehicles is that it doesn’t have a strong reference in our daily lives for respondents to associate it with. Moreover, the respondents have not experienced such technology in any form yet. A web-based survey does not have the required features for a DCE researcher to design choice scenario that best represent the reality and also help the respondents to completely understand them. It is further evident from the high variance in the willingness-to-pay reported in two studies that the web-based survey may not have provided the level of information necessary to respond objectively. McFadden (2015) alluded on this shortcoming in the context of disruptive technologies and suggested the use of immersive simulators that can give the respondents near-real experience. Methodology The core of VIRE is a multimodal agent based microsimulation that simulates pedestrian, cyclists, driver-based vehicles, and automated vehicles on urban networks. The 3D visualization is projected using a gaming engine (Unity3D) to an HMD to give a first person realistic view (see Figure 1 in additional document). Pedestrian flow in the intersections is simulated using Social Force model, while vehicular and cyclist flows on roads are simulated using Car Following models. Voice commands are implemented in the automated vehicle. Signalization in the intersections can be centrally controlled based on the flow optimization. The v2v, v2i, and i2i communication can be simulated and used to propagate the information (e.g. high flow coming from upstream, or accident downstream) and also guide the vehicles dynamically. A physical gaming steering wheel is used to give control to the individual to drive driver-based vehicle on the network. Various levels of automated and manual vehicles can be simulated. The design of choice experiment involves two parts: In the first part the respondent is asked to drive from a familiar origin to destination by controlling the steering wheel in the simulation. They control the car using the gaming steering wheel and view the scene on the HMD. This is performed several times, with a certain temporal distance so that the correlation between each run is minimized. In the second part, the respondent is asked to experience the travel between same origin and destination in an automated vehicle that is receiving live information from other vehicles and infrastructure, and its route is optimized based on the instructions it receives. This is also repeated several times. During these simulations respondent has full 360 o view of their environment in the HMD. At the end of these experiences, the respondent is asked a set of questions regarding the experience, speed, and travel times. The simulation saves the trajectories and a video sensors record the movements of the respondents for further analysis. Based on the data gathered we will develop advanced discrete choice analysis for the acceptability of a system optimum network with highly automated vehicles. Case study A very detailed 3D model of Downtown Montreal has been developed and is used as a case study. References Bansal, P., Kockelman, K.M., Singh, A. (2016) Assessing public opinions of and interest in new vehicle technologies: An Austin perspective, Transportation Research Part C: Emerging Technologies, 67: 1-14. Darbani, J.M., Rezaei, A., Patterson, Z., Zacharias. J. (2013) Video Game vs. Traditional Text-only SP Survey of Neighborhood Choice. In  Proceedings of the 93 rd Annual Meeting of the Transportation Research Board . Washington, DC. Daziano, R.A., Sarrias, M., Leard B. (2016) Are consumers willing to pay to let cars drive for them? Analyzing response to autonomous vehicles. Resources for the Future. RFF DP 16-35. Cherchi, E., Hensher, D.A. (2015) Workshop Synthesis: Stated Preference Surveys and Experimental Design, an Audit of the Journey so far and Future Research Perspectives, Transportation Research Procedia, 11(2015): 154-164. McFadden, D., (2015) Taking Care of Bus-i-ness: Models and Methods for Direct Elicitation of Indirect Preferences. Keynote at 14 th International Conference on Travel Behaviour Research . Windsor, England.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.