Physics-informed machine learning for time-domain inference, at the Moore DDD investigators meeting.