Involvement of artificial intelligence in receiving emergency calls
The subject of the project solution was industrial research focused on the deployment of artificial intelligence for receiving emergency calls during major emergencies using a voice chatbot (voicebot). The result of the project is a functional demonstrator (voicebot), which is deployed on the telephone infrastructure. Voicebot is composed of individual components supplied by project partners.
Problem or opportunity
The consequences of major natural disasters (strong winds, floods, thunderstorms) generate a huge number of emergency calls, and the system is overwhelmed by both the technological part and personnel. The capacity of emergency contact centers is limited and thus emergency lines are not always available. Our goal is to prevent an emergency caller from not being in line, but being dealt with immediately. A possible solution to help overwhelmed operators is the deployment of artificial intelligence to receive emergency calls.
Phonexia is developing technology to transcribe audio to speech or to the text, and this text can then be used by another module to identify what a person needs and for automatic service. Phonexia offers automatic speech transcription, telephony and integration.
- Voicebot backend: voicebot control logic. Its purpose is to receive the call and then process it.
- Speech to Text: technology that recognizes speech in the received recording and then converts it into text.
- Voice biometrics: technology that can identify the caller by voice and determine if he has called before and what his request was.
- We also supply a component of speech technologies called Speech Engine, its purpose is to receive a call from the backend in the form of audio and forward it to the following components for processing.