Real-time systems with a finite (may be big) number of tasks are under consideration. A system is called resolvable if there exists a planning providing in time completion of each task.

The paper deals with the localization problem: on what conditions thef in time completion problem of a system could be reduced to existence of a number of independent and local in time resolvable subsystems.

Full subsystem is defined as a natural local planning object. It is proved that a system is uniquely represented as a union of a number of full subsystems and that a system is resolvable in case all full subsystems are resolvable.