Radar sensing technology is used in the ROGEST project for monitoring purpose. The extraction of the micro-Doppler signature at distinct spatial points of the sensed environment allows to recognize and study in detail the so-called intrinsic motions of the subject under analysis. This technology is useful in several applicative scenarios. Among these, one of the most relevant is the monitoring of the actions performed by workers in environments considered at risk, where movements performed incorrectly, carelessly or under non-optimal emotional states can lead to accidents at work with potentially fatal consequences. The concept of radar prediction was demonstrated by analysing a simple repetitive action (e.g., screwing with a screwdriver) performed by the same worker under different psycho-physical states. The emotional state of the actor is then recognized with the help of algorithms based on artificial intelligence.