“Intelligent transmitters” taking part in distributed and networked control systems are considered. Such sensor systems are able to perform internal data processing, advanced functionalities (e.g. self-diagnoses, online reconfiguration), and to exchange information. A control system made up of cooperating transmitters is therefore presented, using procedures which aim to improve system availability and safety. Taking advantage of the information exchanged between transmitters, two algorithms are proposed in order to i) retain the most confident values for transmitter processing, and ii) perform diagnoses by result comparisons. The control system is modelled by stochastic and coloured Petri nets, and dependability evaluations are performed by Monte Carlo simulations. Availability and safety can be put in balance through the use of the proposed algorithms, and both criteria can be improved under some conditions, notably according to diagnostic coverage of failures.