The study in ViP publication 2015-3 investigates the need and potential for night-time scenarios in driving simulators, determines how such night-time scenarios could be reproduced and identifies the objects most important to reproduce. Although on average 12 out of every 24 hours are dark and considering that most situations are more demanding for drivers in dark conditions, simulations of driving scenarios with different degrees of darkness are not common. The project work comprised a pre-study that involved an investigation of the need and potential of night-time scenarios with the help of input from different stakeholders, consolidation of what is known up to now through benchmarking and state of the art, and a review of available technical solutions. The objective was to identify pros and cons with existing solutions and aspects that are important to consider in order to reproduce the most important components in realistic night-time scenarios. Based on the results, six important use cases were identified and two of these (‘Driver fatigue’ and ‘Objects without light sources’) were studied in more detail. It was concluded that for night-time scenarios there is enough darkness in general in the simulator environment. The question is whether it is possible to create sufficient contrast for objects that are meant to be observable. For daytime scenarios, the light levels in the simulator are clearly unrealistically low and this limitation might even trigger unwanted sleepiness. The use cases identified were very different and there is a need to understand that one solution might not fit all applications. One example is a simulator for testing car headlights, which is a very limited use case with specific requirements compared with a simulator used for behaviour studies, which can employ a simpler, but still high-fidelity, render. Considering ViP’s interest in improving its simulation platform, it is also important to take into account the need for relative and absolute reflection of reality. Whereas it might not be possible to recreate the absolute colour and contrast reflection, it may be easier to create a relative reflection. More research is needed to identify and understand whether a relative model would be sufficient for the simulator. Recommendations for future research fall within three development areas: • Light source simulation • Driving in dark conditions • In-vehicle light level in daylight scenarios. (Author/publisher)
Abstract