Hearing-impaired or distracted vehicle operators may be slow to notice or completely miss alarms from fast approaching emergency vehicles at the risk of impeding the progress of or an actual collision with said emergency vehicle. I present a prototype of a mobile phone application that applies Convolution Neural Network algorithms to distinguish emergency-vehicle generated sonic signatures from other urban sounds and analyzes live sound waves to offer distracted and hearing-impaired drivers augmented information on the classification and directional path of the sound source. All necessary alerts are projected onto a Head-Up Display attached to the vehicle’s windshield, and urban sounds are detected by two Bluetooth stereo microphones. The user may also be notified of nearby emergency vehicles with an optional auditory and tactile alarm.
