The AI-powered machines, capable of identifying patients using facial recognition, can respond to voice commands from staff