REDUCING RISKS FROM EMERGENCY VEHICLE LIGHTS

An emergency vehicle that is displaying emergency lights detects an approaching vehicle, and assesses road conditions, intensity of the emergency lights at the approaching vehicle, a response of the approaching vehicle to the emergency lights, and a state variable related to a driver of the approaching vehicle. The emergency vehicle then obtains a likelihood value that the approaching vehicle will initiate an accident, by supplying information about the road conditions, the intensity of the emergency lights at the approaching vehicle, the response of the approaching vehicle, and the state variable as inputs to a neural network trained to correlate accidents to such information. In response to the likelihood value exceeding a threshold value, the emergency vehicle modifies the emergency lights.