Urban Time-to-Detect Simulator for Vehicle-Developers

The proposed approach starts from an investigation of psychophysical stimuli relevant to the detection of a vehicle in an urban environment. Existing simulators will then be evaluated from the standpoint of their capability of high fidelity reproduction of real stimuli and on their ability to handle virtual stimuli (i.e. from virtual vehicles). Sound Answers will follow an iterative approach to the development of a multi-modal detectability metric, by first conducting experiments of vehicle detection based on one type of high quality cue (as an example, just visual or audio or vibration). Bi-modal detection experiments will then be conducted by using the (one or more) simulators that offer best signal quality for at least two stimuli. The results of both series of experiments will be analyzed to derive a detectability model based on one stimulus only, then a detectability model which accounts for the cross-coupling among two stimuli at a time. Finally, a hypothesis of a multi-modal metric model will be formulated along with a specification document for the simulator to be built in Phase 2.