Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

Prognostics and health management design for rotary machinery systems—Reviews, methodology and applications

Machinery health prognostics: A systematic review from data acquisition to RUL prediction

An overview of time-based and condition-based maintenance in industrial application

There has been a growing interest in monitoring the ongoing "health" of products and systems in order to predict failures and provide warning to avoid catastrophic failure. Here, health is defined as the extent of degradation or deviation from an expected normal condition. While the application of health monitoring, also referred to as prognostics, is well established for assessment of mechanical systems, this is not the case for electronic systems. However, electronic systems are integral to the functionality of most systems today, and their reliability is often critical for system reliability. This paper presents the state-of-practice and the current state-of-research in the area of electronics prognostics and health management. Four current approaches include built-in-test (BIT), use of fuses and canary devices, monitoring and reasoning of failure precursors, and modeling accumulated damage based on measured life-cycle loads. Examples are provided for these different approaches, and the implementation challenges are discussed.

/pdf/prognostics-and-health-management-of-electronics-46ggsuwf7q.pdf

Prognostics and health management of electronics

This paper introduces a method to detect a fault associated with critical components/subsystems of an engineered system. It is required, in this case, to detect the fault condition as early as possible, with specified degree of confidence and a prescribed false alarm rate. Innovative features of the enabling technologies include a Bayesian estimation algorithm called particle filtering, which employs features or condition indicators derived from sensor data in combination with simple models of the system's degrading state to detect a deviation or discrepancy between a baseline (no-fault) distribution and its current counterpart. The scheme requires a fault progression model describing the degrading state of the system in the operation. A generic model based on fatigue analysis is provided and its parameters adaptation is discussed in detail. The scheme provides the probability of abnormal condition and the presence of a fault is confirmed for a given confidence level. The efficacy of the proposed approach is illustrated with data acquired from bearings typically found on aircraft and monitored via a properly instrumented test rig.

A Probabilistic Fault Detection Approach: Application to Bearing Fault Detection

In recent years, numerous machinery health monitoring technologies have been developed by the US Navy to aid in the detection and classification of developing machinery faults for various Naval platforms. Existing Naval condition assessment systems such as ICAS (Integrated Condition Assessment System) employ several fault detection and diagnostic technologies ranging from simple thresholding to rule-based algorithms. However, these technologies have not specifically focused on the ability to predict the future condition (prognostics) of a machine based on the current diagnostic state of the machinery and its available operating and failure history data. An advanced prognostic capability is desired because the ability to forecast this future condition enables a higher level of condition-based maintenance for optimally managing total life cycle costs (LCC). A second issue is that a framework does not exist for "plug-and-play" integration of new diagnostic and prognostic technologies into existing Naval platforms. This paper outlines such prognostic enhancements to diagnostic systems (PEDS) using a generic framework for developing interoperable prognostic "modules". Specific prognostic module examples developed for gas turbine engines and gearbox systems are also provided.

Prognostic enhancements to diagnostic systems for improved condition-based maintenance [military aircraft]

Impact technologies have developed a robust modeling paradigm for actuator fault detection and failure prediction. This model-based approach to prognostics and health management (PHM) applies physical modeling and advanced parametric identification techniques, along with fault detection and failure prediction algorithms, in order to predict the time-to-failure for each of the critical, competitive failure modes within the system. Advanced probabilistic fusion strategies are also leveraged to combine both collaborative and competitive sources of evidence, thus producing more reliable health state information. These algorithms operate only on flight control command/response data. This approach for condition-based maintenance provides reliable early detection of developing faults. As an advantage over 'black-box' health-monitoring schemes, faults and failure modes are traced back to physically meaningful system parameters, providing the maintainer with invaluable diagnostic and prognostic information. The developed model-based reasoner was validated and demonstrated on an electromechanical actuator (EMA) provided by Moog, Inc.

A model-based approach to prognostics and health management for flight control actuators

Disclosed are systems and methods for prognostic health management (PHM) of electronic systems. Such systems and methods present challenges traditionally viewed as either insurmountable or otherwise not worth the cost of pursuit. The systems and methods are directed to the health monitoring and failure prediction of electronic systems, including the diagnostic methods employed to assess current health state and prognostic methods for the prediction of electronic system failures and remaining useful life. The disclosed methodologies include three techniques: (1) use of existing electronic systems data (circuit as a sensor); (2) use of available external measurements as condition indicators and degradation assessor; and (3) performance assessment metrics derived from available external measurements.

Systems and methods for predicting failure of electronic systems and assessing level of degradation and remaining useful life

Actuators are complex electro-hydraulic or mechanical mechanisms utilized in aircraft to drive flight control surfaces, landing gear, cargo doors, and weapon systems. Impact has developed a prognostic and health management (PHM) methodology for these critical systems that includes signal processing and neural network tracking techniques, along with automated reasoning, classification, knowledge fusion, and probabilistic failure mode progression algorithms. The processing utilizes the command/response signal and hydraulic pressure data from the actuators and provides a real-time assessment of the current/future actuator health state. This methodology was applied to F/A-18 stabilator electro-hydraulic servo valves (EHSVs) using test stand data provided by Boeing Phantom works. The automated module demonstrated excellent health state classification results. The prognosis was also successfully performed however no data was available to validate the prediction. These algorithms were developed with consideration to sensor/processing limitations for potential onboard implementation. Many of the PHM elements presented here could also be adapted for other actuator types and applications.

Carl S. Byington

Papers

A Probabilistic Fault Detection Approach: Application to Bearing Fault Detection

Prognostic enhancements to diagnostic systems for improved condition-based maintenance [military aircraft]

A model-based approach to prognostics and health management for flight control actuators

Systems and methods for predicting failure of electronic systems and assessing level of degradation and remaining useful life

Data-driven neural network methodology to remaining life predictions for aircraft actuator components