Repairable systems are found in automobiles, refrigerators, power plants, telephone communication networks, aircraft and defense systems, for example.
A repairable system consists of individual components that age and require maintenance. Repairable systems are involved in many aspects of modern life, including automobiles, refrigerators, power plants, telephone communication networks, aircraft and defense systems. The state of a repairable system changes as its components fail and are subsequently repaired.
The Repairable Systems Simulation (RSS) platform in JMP enables you to explore the reliability within complex repairable systems. Using this platform, you can simulate different system configurations in order to optimize your repairable system. For example, you can extend the useful life of costly components or maximize production outputs while maintaining safe system operation.
We will use an automobile example to illustrate key features in this new and innovative platform in JMP. Some features are available in JMP and alternative software packages. Other features are exclusive to JMP and are designed to equip engineers with more flexibility and command.
The car example in Figure 1 considers four main components in order for an automobile to function. Included are the engine, transmission, battery and tires (all four tires plus the spare).
Figure 1: Automotive Repairable Systems Simulation exampleYou can see that the RSS interface constructs a system much like a traditional Reliability Block Diagram (RBD). You add main components by dragging and dropping various node types from the top panel into the design space. Each component is then connected with arrows that define the system flow and dependencies. Distribution type and parameters can be defined for each block similar to the JMP RBD platform.
Now, since we are interested in managing a complex system that can support a wide variety of maintenance activities, the RSS platform incorporates simulation techniques in order to assess all possible outcomes.
Figure 2: Maintenance actions for automotive engine
In order to prepare for the simulation of possible maintenance outcomes for our design, we need to provide definitions and criteria related to the scenarios and types of maintenance we can perform on each component.
RSS extends the common scenarios of maintenance, which include corrective and preventive maintenance. The typical description of corrective maintenance is maintenance necessary when a component fails. The typical description of preventive maintenance is maintenance which is scheduled to be performed at a specific interval of time. RSS extends the scenarios. In RSS, corrective maintenance is maintenance when any unplanned event happens; preventive maintenance is maintenance which is planned ahead of time.
RSS provides simulation of the common types of maintenance (“replace with new”, “minimal repair”) along with providing many more. For example, RSS supports using refurbished components, changing stress, selectively turning on and off individual components, rescheduling, etc. For user’s convenience, RSS provides some built-in components with behaviors predefined, such as Standby and Stress-sharing blocks.
Figures 1 and 2 show that there are additional colored nodes: orange to specify the type of maintenance “Event” and dark blue to specify the type of “Action” that will take place if an “Event” occurs. These orange and blue maintenance nodes are defined for each main component (light blue node). There are numerous types of Event and Action nodes for the user to choose. Many of these are reflected in the Figures above. Custom options are supported as well, and, depending on the required complexity, Event and Action nodes can be associated from one system component to another.
The final step prior to simulation and analysis is to define the criteria of each Event and Action. You can see examples of this in the right side panel of Figure 2. Next, we define the relationship of the Event and Action nodes by connecting them with drag and drop arrows the same way we connect Main Components. The orange “Event” nodes, dark blue “Action” nodes, and green arrows form a sub-diagram which expresses the maintenance arrangement, which is a visual and easy-to-communicate representation of a long to-do list of “do this when that happens.”
Additionally, you can specify time to complete these actions in many ways, including Triangle or Uniform distributions, Immediate, Constant, etc.
The new RSS Platform User Interface (UI) design introduced in JMP 13 is a unique, state-of-the-art application that allows design engineers, maintenance engineers and system managers to construct complex systems, define component reliabilities and specify maintenance actions in one easy-to-navigate UI. All activity can be accomplished in one innovative work space, so users do not have to navigate a complicated set of hierarchical dialogs that would require continually opening, editing and closing to assess the potential of complex system designs.
Next, as you can see in the upper-right corner of both Figures 1 and 2, you can easily specify simulation parameters, including Time Units or “Clocks.” The times units can also be specified by each component as well.
Lastly, when we run the simulation, JMP generates a data table that includes predicted actions, events and time lines, etc., that we expect to see under use conditions. Figure 3 represents some examples of the Results Explorer that is generated automatically from the simulation data table.
Figure 3: Sample report from Repairable Systems Simulation platformLet us know what you think of this new capability in JMP.
Note: My colleague @peng_liu in JMP Development co-authored this post with me.