TwinCAT 3 integrates machine learning (ML) software, giving ML applications the same advantages of system openness familiar from PC-based control technology, and supports real-time ML to enhance machine performance, for example, through predictive maintenance, process self-optimization and autonomous detection of process anomalies. Controls engineers no longer have to design solutions for specific tasks and turn those solutions into algorithms—instead, the controller can learn the desired algorithms from exemplary process data.