Prognostics is an engineering discipline focused on predicting the time at which a system or a component will no longer perform its intended function. This lack of performance is most often a failure beyond which the system can no longer be used to meet desired performance. The predicted time then becomes the remaining useful life (RUL), which is an important concept in decision making for contingency mitigation. Prognostics predicts the future performance of a component by assessing the extent of deviation or degradation of a system from its expected normal operating conditions. The science of prognostics is based on the analysis of failure modes, detection of early signs of wear and aging, and fault conditions. An effective prognostics solution is implemented when there is sound knowledge of the failure mechanisms that are likely to cause the degradations leading to eventual failures in the system. It is therefore necessary to have initial information on the possible failures (including the site, mode, cause and mechanism) in a product. Such knowledge is important to identify the system parameters that are to be monitored. Potential uses for prognostics is in condition-based maintenance. The discipline that links studies of failure mechanisms to system lifecycle management is often referred to as prognostics and health management (PHM), sometimes also system health management (SHM) or--in transportation applications--vehicle health management (VHM) or engine health management (EHM). Technical approaches to building models in prognostics can be categorized broadly into data-driven approaches, model-based approaches, and hybrid approaches.
Data-driven prognostics usually use pattern recognition and machine learning techniques to detect changes in system states. The classical data-driven methods for nonlinear system prediction include the use of stochastic models such as the autoregressive (AR) model, the threshold AR model, the bilinear model, the projection pursuit, the multivariate adaptive regression splines, and the Volterra series expansion. Since the last decade, more interests in data-driven system state forecasting have been focused on the use of flexible models such as various types of neural networks (NNs) and neural fuzzy (NF) systems. Data-driven approaches are appropriate when the understanding of first principles of system operation is not comprehensive or when the system is sufficiently complex such that developing an accurate model is prohibitively expensive. Therefore, the principal advantages to data driven approaches is that they can often be deployed quicker and cheaper compared to other approaches, and that they can provide system-wide coverage (cf. physics-based models, which can be quite narrow in scope). The main disadvantage is that data driven approaches may have wider confidence intervals than other approaches and that they require a substantial amount of data for training. Data-driven approaches can be further subcategorized into fleet-based statistics and sensor-based conditioning. In addition, data-driven techniques also subsume cycle-counting techniques that may include domain knowledge.
The two basic data-driven strategies involve (1) modeling cumulative damage (or, equivalently, health) and then extrapolating out to a damage (or health) threshold, or (2) learning directly from data the remaining useful life.
As mentioned, a principal bottleneck is the difficulty in obtaining run-to-failure data, in particular for new systems, since running systems to failure can be a lengthy and rather costly process. When future usage is not the same as in the past (as with most non-stationary systems), collecting data that includes all possible future usages (both load and environmental conditions) becomes often nearly impossible. Even where data exist, the efficacy of data-driven approaches is not only dependent on the quantity but also on the quality of system operational data. These data sources may include temperature, pressure, oil debris, currents, voltages, power, vibration and acoustic signal, spectrometric data as well as calibration and calorimetric data. Features must be extracted from (more often than not) noisy, high-dimensional data.
Selecting proper prognostics algorithm for each application is a challenging factor in applying data driven prognostics methods. In a survey article, authors have prepared useful information to conclude pros and cons of various prognostics algorithms in fault diagnosis and failure prediction of rotating machineries.
Model-based prognostics attempts to incorporate physical understanding (physical models) of the system into the estimation of remaining useful life (RUL). Modeling physics can be accomplished at different levels, for example, micro and macro levels. At the micro level (also called material level), physical models are embodied by series of dynamic equations that define relationships, at a given time or load cycle, between damage (or degradation) of a system/component and environmental and operational conditions under which the system/component are operated. The micro-level models are often referred as damage propagation model. For example, Yu and Harris's fatigue life model for ball bearings, which relates the fatigue life of a bearing to the induced stress, Paris and Erdogan's crack growth model, and stochastic defect-propagation model are other examples of micro-level models. Since measurements of critical damage properties (such as stress or strain of a mechanical component) are rarely available, sensed system parameters have to be used to infer the stress/strain values. Micro-level models need to account in the uncertainty management the assumptions and simplifications, which may pose significant limitations of that approach.
Macro-level models are the mathematical model at system level, which defines the relationship among system input variables, system state variables, and system measures variables/outputs where the model is often a somewhat simplified representation of the system, for example a lumped parameter model. The trade-off is increased coverage with possibly reducing accuracy of a particular degradation mode. Where this trade-off is permissible, faster prototyping may be the result. However, where systems are complex (e.g., a gas turbine engine), even a macro-level model may be a rather time-consuming and labor-intensive process. As a result, macro-level models may not be available in detail for all subsystems. The resulting simplifications need to be accounted for by the uncertainty management.
Hybrid approaches attempt to leverage the strength from both data-driven approaches as well as model-based approaches. In reality, it is rare that the fielded approaches are completely either purely data-driven or purely model-based. More often than not, model-based approaches include some aspects of data-driven approaches and data-driven approaches glean available information from models. An example for the former would be where model parameters are tuned using field data. An example for the latter is when the set-point, bias, or normalization factor for a data-driven approach is given by models. Hybrid approaches can be categorized broadly into two categories, 1) Pre-estimate fusion and 2.) Post-estimate fusion.
The motivation for pre-estimate aggregation may be that no ground truth data are available. This may occur in situations where diagnostics does a good job in detecting faults that are resolved (through maintenance) before system failure occurs. Therefore, there are hardly any run-to-failure data. However, there is incentive to know better when a system would fail to better leverage the remaining useful life while at the same time avoiding unscheduled maintenance (unscheduled maintenance is typically more costly than scheduled maintenance and results in system downtime). Garga et al. [REF] describe conceptually a pre-estimate aggregation hybrid approach where domain knowledge is used to change the structure of a neural network, thus resulting in a more parsimonius representation of the network. Another way to accomplish the pre-estimate aggregation is by a combined off-line process and on-line process: In the off-line mode, one can use a physics-based simulation model to understand the relationships of sensor response to fault state; In the on-line mode, one can use data to identify current damage state, then track the data to characterize damage propagation, and finally apply an individualized data-driven propagation model for remaining life prediction.
Motivation for post-estimate fusion is often consideration of uncertainty management. That is, the post-estimate fusion helps to narrow the uncertainty intervals of data-driven or model-based approaches. At the same time, the accuracy improves. The underlying notion is that multiple information sources can help to improve performance of an estimator. This principle has been successfully applied within the context of classifier fusion where the output of multiple classifiers is used to arrive at a better result than any classifier alone. Within the context of prognostics, fusion can be accomplished by employing quality assessments that are assigned to the individual estimators based on a variety of inputs, for example heuristics, a priori known performance, prediction horizon, or robustness of the prediction.
Prognostic performance evaluation is of key importance for a successful PHM system deployment. The early lack of standardized methods for performance evaluation and benchmark data-sets resulted in reliance on conventional performance metrics borrowed from statistics. Those metrics were primarily accuracy and precision based where performance is evaluated against actual End of Life (EoL) typically known a priori in an offline setting. More recently, efforts towards maturing prognostics technology has put a significant focus on standardizing prognostic methods, including those of performance assessment. A key aspect, missing from the conventional metrics, is the capability to track performance with time. This is important because prognostics is a dynamic process where predictions get updated with an appropriate frequency as more observation data become available from an operational system. Similarly, the performance of prediction changes with time that must be tracked and quantified. Another aspect that makes this process different in a PHM context is the time value of a RUL prediction. As a system approaches failure, the time window to take a corrective action gets shorter and consequently the accuracy of predictions becomes more critical for decision making. Finally, randomness and noise in the process, measurements, and prediction models are unavoidable and hence prognostics inevitably involves uncertainty in its estimates. A robust prognostics performance evaluation must incorporate the effects of this uncertainty.
Several prognostics performance metrics have evolved with consideration of these issues:
A visual representation of these metrics can be used to depict prognostic performance over a long time horizon.
The industrial applications of PHM are quite diverse in terms of industry, with examples found in manufacturing, automotive, heavy industry, aerospace, power generation, and transportation. With respect to manufacturing, there has been considerable work for rotating machinery, including PHM development and solutions for the machine tool industry. Examples include methods and software for monitoring spindle bearing health based on vibration and motor current, a cloud based monitoring architecture for relating tool wear health to part quality, and numerous works on monitoring the health condition of a machine tool feed axis. A low cost and practical method for monitoring the machine tool feed axis was demonstrated in a production environment, in which only controller signals were used to detect the early symptoms of pulley degradation prior two weeks before the pulley axis failed. For automotive manufacturing, there has recent developments on monitoring the health condition of industrial robots, using available controller signals such as motor current; this approach represents a practical approach for monitoring a fleet of industrial robots. Methods based on frequency analysis and classification algorithms for detecting the early symptoms of surge for air compressors have also been successfully demonstrated and implemented for an automotive manufacturing plant. Data mining and advanced analytical approaches have also been developed for continuous manufacturing production lines and semiconductor manufacturing applications. In terms of success in the manufacturing sector for PHM solutions, there is some economic numbers that can be reported. For example, the National Science Foundation funded in independent economic impacts study on Industry/University Cooperative Research Centers (I/UCRC) and surveyed 5 industrial members of the Center for Intelligent Maintenance Systems; the 5 companies (predominantly manufacturing applications) reported a savings of over $855 Million U.S dollars based on the successful implementation of the predictive monitoring and PHM solutions.
Heavy vehicles used in construction, agriculture, and mining, are also seeing greater interest in predictive monitoring and PHM technology. Original equipment manufacturers of these heavy duty vehicles, such as Komatsu and Caterpillar already have the infrastructure in place for remote monitoring, and are now developing the advanced data analysis algorithms to detect the vehicle problems at an early stage. Original equipment manufacturers for underground mining are also developing the necessary infrastructure and analysis algorithms with the idea of providing similar PHM service solutions.
Commercial implementations of PHM solutions in the power generation industry are also increasing, with applications that focus on rotating machinery and turbines, to early problem detection algorithms based on data from the supervisory control data acquisition (SCADA) system. In addition, to some of the larger assets, vibration monitoring and intelligent analytics are also being considered for the balance of plant (fans, pumps) equipment used in power generation.
Renewable energy applications, such as wind turbines, are also an industry sector that has received considerable attention regarding PHM technology and commercial solutions. Approaches based on using statistical methods for modeling the normal relationship between input parameters such as wind speed and output parameters such as generator power have been used to successfully monitor the wind turbines performance degradation. Wind turbine drivetrain condition monitoring solutions based on vibration data have also seen considerable research work and some commercial monitoring products are also available. The wind turbine drive train monitoring represents a more challenging PHM application, due to the rotational speed fluctuations and low rotational speeds for the input shaft, time-varying load conditions, and the more complicated vibration algorithms needed for monitoring planetary gearbox health.
The aerospace and defense sector has several research studies in the PHM area and some fielded systems that have some level of PHM functionality. The health and usage management systems (HUMS) is an example of a fielded PHM solution for rotorcraft, that can detect several different types of problems using vibration and other signals, from shaft unbalance, to gear and bearing deterioration. In addition, it is reported that the HUMS system provided significant maintenance cost reductions and improved fleet availability when compared with rotorcraft units that did not have the HUMS systems. Aircraft engines are another application in which there is considerable PHM technology that is being used and developed. Original equipment manufacturers, such as General Electric Aviation have monitored aircraft engines for over 15 years and are providing diagnostic services to detect the early symptoms of engine problems before they lead to operational downtime. In terms of research and development efforts, the Joint Strike Fighter program allocated significant resources for PHM development and implementation. PHM research and case studies for military ground vehicles are also being conducted for engines, alternator, and structural components; however it seems that there are more fielded systems for the aerospace platforms at this time.
There are numerous research studies in the automotive sector that aim at providing a more advanced monitoring functionality for monitoring key vehicle components, such as vehicle batteries, vehicle alternators, and internal combustion engines. Methods based on looking at unusual patterns for a particular vehicle when compared to the fleet have also showed promise in a research setting. Electric vehicle and battery health monitoring and prognostic work has also seen an increasing level of research and development. Algorithms for estimating the electric vehicles battery state of charge and state of health have been successfully demonstrated in several research studies. A more challenging problem is predicting the remaining driving range for an electric vehicle, since this depends on not just the battery state of charge, but also environmental factors, the road and traffic conditions, driving behavior, among other factors.
Monitoring the condition of the rolling stock and railway infrastructure has also been an area that has received considerable attention. Monitoring the condition of the track infrastructure using vibration, displacement, and other measurements has been conducted in several research studies, with some initial systems being implemented. For the track infrastructure, vibration data analysis based on the magnitude, wavelength and time-frequency characteristics along with statistical or pattern recognition tools can be used to assess the track condition with respect to corrugation, rolling contact fatigue defects, and geometry and alignment issues. Point Machines (devices used to operate railway turnouts), are also a target for PHM technology, in which the electrical signals and statistical or pattern recognition analysis methods can be used to catch the early symptoms of point machine degradation prior to failure. Commercial solutions for rolling stock condition monitoring are also being provided by original equipment manufacturers. An example commercial offering is the TrainTracer product, which provides real-time data collection and remote monitoring of rolling stock systems and components.
For most PHM industrial applications, commercial off the shelf data acquisition hardware and sensors are normally the most practical and common. Example commercial vendors for data acquisition hardware include National Instruments and Advantech Webaccess; however, for certain applications, the hardware can be customized or ruggedized as needed. Common sensor types for PHM applications include accelerometers, temperature, pressure, measurements of rotational speed using encoders or tachometers, electrical measurements of voltage and current, acoustic emission, load cells for force measurements, and displacement or position measurements. There are numerous sensor vendors for those measurement types, with some having a specific product line that is more suited for condition monitoring and PHM applications.
The data analysis algorithms and pattern recognition technology are now being offered in some commercial software platforms or as part of a packaged software solution. National Instruments currently has a trial version (with a commercial release in the upcoming year) of the Watchdog Agent® prognostic toolkit, which is a collection of data-driven PHM algorithms that were developed by the Center for Intelligent Maintenance Systems. This collection of over 20 tools allows one to configure and customize the algorithms for signature extraction, anomaly detection, health assessment, failure diagnosis, and failure prediction for a given application as needed. Customized predictive monitoring commercial solutions using the Watchdog Agent toolkit are now being offered by a recent start-up company called Predictronics Corporation in which the founders were instrumental in the development and application of this PHM technology at the Center for Intelligent Maintenance Systems. Other commercial software offerings focus on a few tools for anomaly detection and fault diagnosis, and are typically offered as a package solution instead of a toolkit offering. Example includes Smart Signals anomaly detection analytical method, based on auto-associative type models (similarity based modeling) that look for changes in the nominal correlation relationship in the signals, calculates residuals between expected and actual performance, and then performs hypothesis testing on the residual signals (sequential probability ratio test). Similar types of analysis methods are also offered by Expert Microsystems, which uses a similar auto-associative kernel method for calculating residuals, and has other modules for diagnosis and prediction.