Prediction diagnostic for smart cities systems

Systems applied in the last few years for an improvement of the operation of tools ensuring the most critical functions in large cities must provide their functions exploiting various methods from the area of advanced informatics and system theory. One of the most important of them is the prediction diagnostics being a tool allowing how long particular system (or the whole smart cities system alliance) can still operate well, or when it approaches to the end of such reliable state. In this contribution the possibilities of prediction diagnostic apparatus are discussed as well as limitations coming from the fact that these systems must be very often considered as uncertain, especially if they interact with human factor.