A Condition Based Maintenance Plus (CBM+) strategy can help designers in military programs increase reliability and availability, improve maintenance practices, and lower life cycle costs. Low-cost line replaceable units (LRUs) designed for condition monitoring and health assessment can be a large part of a CBM+ integrated strategy.

The Condition Based Maintenance Plus DoD Guidebook [published in March 2017 by the Department of Defense (DoD)] describes CBM+ as a “conscious effort to shift equipment maintenance from an unscheduled, reactive approach at the time of failure to a more proactive and predictive approach that is driven by condition sensing and integrated, analysis-based decisions.” CBM+ implementation is usually seen in high-priced equipment, or in equipment with the most critical failure-effect modes. According to the DoD guidebook, operations and support (O&S) costs account for 65 to 80 percent of a program’s total lifecycle cost. Lower-cost LRUs generally do not feature the condition monitoring required to feed sufficient prognostic data to CBM+ processing. Such LRUs simply provide post-fault visibility into the reactive maintenance system, thereby limiting their usefulness.

As CBM+ continues to gain traction, the benefits to the military will be increasingly dependent on the level of participation of the downstream subsystem diagnostics. A complete reliability-centered maintenance (RCM) analysis that would provide a full set of rules for predictive fault assessment is uncommon in smaller development contracts.

Even using lower-cost, lower-utility devices, it is possible to implement monitoring of critical device status using common components in a way that will enable products to be in a position to participate in CBM+. A well-instrumented LRU, even with less sophisticated health monitoring, can provide valuable insight into unit operation and move a device closer to proactive maintenance.

Figure 1: CBM+ infrastructure areas.

Architecting for condition monitoring

Figure 1 illustrates the CBM+ infrastructure areas. Compliant LRUs in this environment typically satisfy requirements in the areas of sensors, condition monitoring, health assessment, communications, and human interfaces. Design features that perform processing in these areas can be implemented with reasonably low recurring costs. Although the initial development cost is incrementally higher than most common go/no-go diagnostics, the high reuse, longer service life, and potentially lower average maintenance cost make a compelling business case.

Robust health assessment depends on detailed rules from RCM analysis, but at the LRU level, a reliability prediction, along with a failure modes and effects criticality analysis (FMECA), provides enough information to develop meaningful health assessment algorithms. Such monitoring and analysis can uncover trends such as increased current flow, ripple on DC power lines, or a compromised heat map. If health assessment still proves too costly, maintaining a log for future analysis is a reasonable compromise. Valuable data can be extracted to determine if a unit back for repair experienced a shock, vibration, temperature, or input power event.

In a vehicular or avionics display unit (DU), for example, typical diagnostics include memory, communications port internal loopback, power fail, and possibly LED [light-emitting diode] driver voltage monitoring. If there is a single-board computer (SBC) that features diagnostics middleware, additional measured power, temperature, and power cycle count data may be available. Built-in test (BIT) options include power up, background, initiated, and possibly one or more operator-involved visual tests.

A DU architected for condition monitoring is shown in Figure 2. Temperature, power, acceleration, and light sensors measure and transmit key operating parameters to the microcontroller for subsequent health assessment processing.

Figure 2: Display unit architected for condition monitoring (Image courtesy IEE).

Temperature sensors

The low cost of thermal sensors enables the developer to use multiple sensors to gather temperature data from many locations to produce a high-resolution heat map. Candidate designs range from higher-priced sensors with built-in calibration and serial communication, down to very low-cost thermistors that require additional analog multiplexing to provide the signal to the MCU. The stylized depiction in Figure 3, the output of a thermal-simulation study, is illustrative of a typical thermal profile of a circuit card assembly. When validated with empirical data, the design engineer can readily select a few key locations for thermal measurements to support a real-time measurement of the circuit card thermal profile. With the aid of RCM+, this temperature sensor mapping can provide effective measurement of the card and support detection of the onset of failure conditions.

Figure 3: Circuit board assembly heat map visualization and actual measurement (Image courtesy IEE).

Power sensors

Voltage and current monitoring of primary and secondary power have been simplified with the availability of low-cost monolithic devices. Changes in voltage and current that do not correlate to temperature, CPU utilization, or other factors are an indication of the onset of a potential failure.

Luminance sensors

The display unit can track the backlight intensity, ambient light intensity, backlight drive current and voltage, temperature, and elapsed time to piece together trend data on display degradation. Luminance detection may be used to increase fault detection in lieu of operator feedback. Although a display failure may be obvious to the operator, some system designs to not allow operator interaction to be part of fault detection and will not factor into BIT verticality analysis.

Acceleration sensors

Accelerometers can detect shock and vibration events, and can warn that the LRU has been subjected to them, for failure prediction or post-failure root cause analysis. Accelerometers produce large amounts of data, so their output should be processed through a field-programmable gate array (FPGA) or other data-acquisition preprocessor to facilitate storing only the acceleration events that meet threshold criteria. More sophisticated designs may incorporate enough power holdup capacity to detect removal and post-removal handling events during which the LRU is particularly vulnerable.

Health monitoring

The integrated diagnostics engineer must determine what to do with all this data. The absence of complete systemwide RCM analyses leaves the health assessment processing to the supplier, who should perform sufficient failure analysis to provide basic rules for health monitoring. Many failure thresholds are specific to devices and require a large amount of specification details obtained from component data sheets. Although it may take time and effort to provide meaningful LRU machine health data into the CBM+ environment, health monitoring is a critical component for full participation.

Communications

In our example, the display unit is responsible for providing pertinent operating parameters and health status data to the mission computer. The system is responsible for making the data accessible to all levels of the CBM+ environment.

Human interfaces

The DU presents critical operating parameters and health assessment data to the operator for workaround actions during mission conduct. This data, in the more detailed form of the fault log, lets the maintenance operator use the CBM+ Communications infrastructure to accomplish effective troubleshooting and repair. The unit also enables the mission and maintenance operators to initiate diagnostics and visual tests that require operator feedback.

Sensors the key

Lower-cost LRUs can participate more fully in a CBM+ environment by implementing sensors, condition monitoring, health assessment, communications, and human interface functions. If all such units implement these functions to some degree, designers of even critical equipment used by the military will find a significant impact on total life cycle costs and reliability.

John Rodwig is a Director of Program Management at Industrial Electronic Engineers, Inc (IEE). He has more than 30 years of technical and management experience in defense and telecommunications, and coholds a patent for a radar scan converter. John earned a BS in electrical engineering from Tulane University and an MS in engineering management from California State University, Northridge. He can be reached at [email protected].

IEE ieeinc.com