A Markov tool is both is a drawing and calculation tool. It is used to create and evaluate Markov diagrams, which are visual representations of the system flow between the possible states of a system. A Markov diagram can be as simple or complex as you wish. Building a Markov diagram consists of inserting the good, failed, and degraded states of the system.

Once you have inserted all possible system states in the diagram, you assign calculation properties to them. You then connect the states. Each state can have zero, one, or more transitions, which are connections to other states. Zero transitions from a state indicate system failure, where no further transitions need to be considered. Multiple transitions from a state indicate that multiple failure modes correspond to that state.

You generally begin with an initial state where all components are operational (good) and end with a state where all components have become totally inoperable (failed). The initial state, which is marked by a bold arrow, considers all components and their failure modes in sequence. The transitions (arcs) that you draw to join states end at the state in which the system is totally inoperable.

To indicate that a component has failed, you draw a transition to the next system state (failed) and assign a failure rate in its calculation properties. To indicate that a failed component has been repaired, you draw a transition back to the original system state (good) and assign a repair rate in its calculation properties. If a transition is already present between two states, then you can add the new transition rate to the existing one.

After failure and repair data is specified for the states and transitions in the Markov diagram, the Markov calculation engine can calculate many different reliability measures, including reliability, availability, MTBF, and more. If you later change the configuration of the states or transitions, when the Markov diagram is recalculated, the results change. Thus, a Markov tool provides an efficient and effective way to compare various configurations to find the best overall system design. The calculation results for this modified Markov diagram then show how much system reliability and availability have improved.