Condition detection in HVAC systems is becoming increasingly important in order to recognize faults and inefficient operation at an early stage. This article aims to illustrate a short-term analysis system and demonstrate its performance in practical operation. The developed modular measurement system works without mechanical intervention in the plant system. It’s a so-called non-invasive system. That means, that the measurement system can be installed during the normal operation process. A special installer is not necessary. The system is supposed to take the data of an existing device (e.g., gas boiler or heat pump) and record them. With this data it is possible to analyze the HVAC System. On the one hand, it is possible to monitor them, calculate key figures, forecasts and forward the data to a central processing unit. One the other hand, it is possible to analyze the data with different AI-methods for an automatic fault detection of the building and the plant system. The fault detection includes e.g. the control of the volume flow, the flow temperature (measured vs. set), operating hours and limit values for electrical power consumption or of the heat supply. As a result, it is possible to model the thermal/electrical load of a building and evaluate the system. This will be done as part of a short-term measurement with a planned duration of three to four weeks, depending on the weather conditions.