Operation of a magnetic fusion experiment, such as JET, relies on the availability of real-time (RT) control schemes, which supervise the plasma as it approaches the expected target performance while maintaining the integrity of the machine and its subsystems. At JET, there have been a series of recent efforts since (Lennholm M. et al 2017 Fusion Engineering and Design 123 535?540) to develop and test RT control schemes in preparation for the upcoming Deuterium-Tritium (DT) campaign. When operating JET in DT, each plasma discharge will in fact be a precious resource, being both T and neutron budget limited. Among the developed control schemes, this paper deals with the isotope ratio controller, which will maintain the required 50:50 DT ratio needed to favor nuclear fusion processes; the dud detector (L. Piron et al 2019 Fusion Engineering and Design 146 1364?1368), which will terminate a discharge moving toward controllers for detecting excessive radiation. Moreover, brandnew detectors, also based on machine learning approaches, have been implemented for detecting off-normal events or pre-disruptive states and have been included in the Plasma Event TRiggering and Alarms system (C. I. Stuart et al 2020 SOFT conference). Work is also ongoing to deploy into JET the RAPTOR suite, a RT observer for plasma state monitoring (C. Piron et al 2020 SOFT conference), and to identify control schemes within RAPTOR capabilities, which could contribute to support the development of high performance plasma scenarios.