This work consideres a geographically localized computerized control sustem. The aim is to determine a policy of tolerance to transient faults, a well as a basic architecture for the control system. A discussion of the allocated effort to implement it is included here. They included the characterization of physical and human faults, masking, detection and recovery of physical faults. The concept of distinctness is considered as an alternative to detection and toleranc3e to human faults and also with respect to the capability to provide an effective detection of physical faults of common mode. This approach provides an independing among redundant modules when the same fault affects them in an identical way. This work presents a basic architecture which illustrates the use of the concept of distinctness, through the emulation of a microprocessor.