Motivation Introduction to PHM Objectives of PHM:

Availability
Reliability
Safety
Overview of maintenance strategies
All technical systems will eventually fail
Unforeseen failures lead to:
Expensive downtime
High maintenance costs
Catastrophic results
Prognostics and Health Management (PHM)
Definition
Capability to contain, prevent, detect, diagnose, predict, respond to, and recover from conditions
that may interfere with nominal system operations.
Prognostics is the process of predicting the future reliability of a product by assessing the extent
of deviation or degradation of a product from its expected normal operating conditions.
Health monitoring is a process of measuring and recording the extent of deviation and
degradation from a normal operating condition.
PHM involves prognostics and decision making based on prognostic results
Consumers:
Challenges

Examples of fault progression are difficult to find due to periodic maintenance and component
replacement
Sensor noise makes it hard to distinguish small, gradual deviations in performance
Limited sensor sets
Sparse set of sensors
Limited information from sensors
Discrete open/closed sensors
Component/System deterioration
Many failure mechanisms can be traced to an underlying deterioration process
Deterioration eventually leads to weakness that can cause failure

Deterioration can be represented or modelled in form of a curve

Y-axis deterioration (damage index/health index)
X-axis representing units of time (cycles/revolutions/odometer readings)
The rate of deterioration process can be modelled as either increasing or decreasing

Most deterioration mechanisms are of increasing type with convex shape, e.g. fatigue crack
growth
Deterioration can also be directly measured (e.g. wear) or derived from measurements (e.g. HI
from vibration measurements)
Objective of PHM

To increase
Reliability
Probability that a unit/system is functioning as required under specified operating conditions at a
given time.
Availability
The ability of a unit/system to be functional when it is required for use or operation
Safety
The ability of a unit/system not to harm the environment, human beings, animals or neighboring
equipment.
AVAILABILITY
Availability is the ability of a service or a system to be functional when it is required for use or
operation;
it is a function of the system’s reliability and maintainability

Reliability is the probability that an item will not fail

Maintainability is the probability that a failed item can be successfully restored to operation

Reliability or Maintainability
Availability

Uptime: accumulated uptimes (system is up and running when requested for operation)
Downtime: accumulated downtimes (system is down undergoing a repair, replacement or waiting
for spares, while requested for operation)
TYPES OF AVAILABILITY
Different types of availability can be measured. These are generally, divided into the following
types of measures:
Time-Interval Based Availability Measures
Instantaneous (or point availability) – probability that an item will be able to perform its required
function at the instant it is required (e.g., Average uptime, Steady-state)
Downtime Based Availability Measures
Availability measures that are defined in terms of what activities are included in the downtime
(e.g., Inherent, Achieved, Operational)
Mission-Oriented Availability
Probability that each individual failure occurring in a mission of a specific total operating time
can be repaired in a time that is less than or equal to some specified time length.
Material Availability
Percentage of the total inventory of a system operationally capable (ready for tasking) or
performing an assigned mission at a given time, based on materiel condition.
Design for availability
To support availability contracts you need a “design for availability” capability
Design for Availability: Determine system parameters that satisfy a specific availability
requirement
 Availability Optimization – seeks to determine design parameters that maximize availability for
“simplified” systems

Only provide solution at a single selected points in time (not at all times)
Implicitly assume that all uptimes are the same and all downtimes are the same (simplifying
assumptions make these solutions non-applicable to real problems)
Don’t seek to “meet” a minimum, rather they attempt to maximize
AVAILABILITY CALCULATIONS
Discrete Event Simulation – create timelines of a set of events and solve for availability
(alternatives: Markov models and Petri-nets)
General, all uptimes can differ, all downtimes can differ
Availability can be evaluated at all points in time
Discrete event simulators only run in the forward direction (increasing time)
Traditionally, availability is only an output (never an input)
If you want to meet an availability requirement you must iterate – which becomes virtually
impossible when uncertainties are introduced
Overview of Maintenance strategies
Maintenance strategies:
Corrective maintenance
Preventive maintenance
Reliability centered maintenance (RCM)
Condition based maintenance (CBM)
Proactive maintenance
Designing to prevent failure
Works on root causes of failure

Evolution of maintenance strategies

Condition Based Maintenance (CBM)
Condition monitoring through sensors/data acqusition
Fault diagnostics – fault detection, isolation and identification
Prognostics – estimating when the system is likely to fail (estimating remaining useful lifetime
(RUL))

Chapter 3: Prognostics (e-maintenance)

Introduction
Prognostics is derived from prognosis which is the process associated with the proactive
maintenance where there is a prediction of the future state of a system or a component and
thereby informs the measures which can be taken to prevent occurrence of failure. Prognosis can
as well be defined as the ability to predict the future condition of a machine based on the current
diagnostic state and the available operating and failure history data (Byington, Roemer and
Galie, 2002).
This process is developed from the principle of proactivity which is initiated in condition based
maintenance (CBM) and in E-maintenance. E-maintenance is the concept of maintenance
management whereby the various components of the system are monitored and managed over
the internet. This system enhances maintenance efficiency as well as presenting an opportunity to
optimize maintenance related workflow.
Technology involved
In this system, there is interaction and collaboration with other processes by exchanging and
sharing information. The structure of the open system often employed is shown in figure 1
below;
Figure 1: Open system architecture (Lebold and Thurston, 2001)
Presentation layer is the man-machine interface which may question the other layers. Decision
support utilizes inventory of spares, manning and logistics among others so as to assemble
maintenance options. Prognostics considers component health assessments to predict future
health. Health assessments assesses historical data to determine current health. Condition based
monitoring gathers specific data and makes a comparison with the specific predetermined values.
Signal processing provides low level computation on sensor data. Data acquisition involves
conversion and formatting of analogue output from transducers to digital data. Transducers
converts the stimuli to electrical signal which can be used in the system (Lebold and Thurston,
2001).
E-maintenance is often considered as a software platform which makes use of the internet for
monitoring functions. Some of the commonly used e-maintenance platforms are PROTEUS,
TELMA, ENIGMA, DYNAweb, SEMATECH among others (Campos, J. and Prakash, 2006). In
this system, there is local real time maintenance system and centralized maintenance system with
some remote stations. This makes it possible to monitor the performance of a system which is
very far from the station. Figure 2 below shows the integration of all components of maintenance
in E-maintenance using TELMA platform.
Figure 2: Integration of maintenance components in E-maintenance (Levrat, 2014)

Vibration analysis
This is a process of measuring the level of vibration and frequency in a machine. The
information obtained can then be used to analyze the level of performance and health of
components and other related machines. With vibration analysis, some of the problems which
can be detected include imbalance, mechanical looseness, misalignment, bearing failures, natural
and resonance frequencies, bent shafts, electrical motor faults, gearbox failures, critical speeds
and cavitation in pumps.
Vibration measurement is the most common condition monitoring data for dynamic systems,
especially rotating machinery.
Faults lead to a change in the mass-stiffness-damping characteristics of a system (Factors that
influence vibrations)
The vibrations are recorded for a specified period of time at predetermined intervals depending
on typical lifetime or expected lifetime of the system.

Data analysis and interpretation

The analysis of data related to vibration takes into consideration some parameters which are
characteristic of the waves formed. These are;
Amplitude – this is the maximum extension of the oscillation and it is measured from the lowest
point to the highest point of the waveform. It is related to the amount of movement. The RMS
value (Root Mean Square) describes the amount of energy contained in this vibration. RMS is
the most used parameter to measure the intensity of the vibration. It helps in detecting the
severity of the fault condition or defect.
Frequency – It measures rate at which movements in vibration occur per second Hz. It helps in
indicating the cause if the defect.
Phase – It is a relative time difference between two signals measured in units of angle as opposed
to time. It only works if the two signals being compared are of the same frequency. It helps in
determining the cause of the defect.
The measurement parameters for vibration analysis are (Scheffer, C. and Girdhar, 2004);
Acceleration: It places greater importance on high frequencies. An acceleration signal is not
exclusive, it can be converted to velocity or displacement.
Displacement: It looks at low frequencies. Used when examining the broad picture of mechanical
vibrations. It can be used to discover unbalance in a rotating part due to a significant amount of
displacement at the rotational frequencies of the machine shaft.
Velocity: It is related to the destructive force of vibration which makes it the most important
parameter. It places equal importance on both high and low frequencies. The RMS value of
velocity (measured in the range of 10 to 10,000 Hz) shows the best sign of vibration severity.
RMS is calculated by multiplying peak amplitude by 0.707.
Figure 3 shows the graphs of displacement, velocity and acceleration while figure 4 shows the
peak and RMS amplitude values.
Figure 3: Displacement, velocity and acceleration graphs (Scheffer, C. and Girdhar, 2004)

Figure 4: Peak and RMS values (Scheffer, C. and Girdhar, 2004)

Steps;
Quantify the level of vibration
In data analysis, the first step is to measure and quantify the level of vibration. It can be
determined by the amplitude of vibration which describes the severity. Peak-to-peak values
indicate the maximum excursion of the wave which is a useful quantity.
Time waveform analysis
Time waveform is the raw data generally acquired by the analyzer and also plays a vital role in
the analysis process. The amplitude units and the acceleration peak to peak values will be the
initial focus for the analyst.
The amplitude reflects the severity of the vibration against time domain (Scheffer, C. and
Girdhar, 2004). This is also known as the machine raw data, as it vibrates while running. The
time waveform contains very important information as it is the source that is used to generate the
spectral data and trend data that are commonly used for fault detection.
Fast Fourier transformation (FFT)
This makes it possible to take vibration time waveform obtained above and mathematically
calculate vibration frequencies and amplitude(Luo, S., Cheng, J. and Fu, 2015). The procedure of
doing transformation is spectrum analysis which is done with the help of FFT analyser. The
frequencies obtained at a given instance is characteristic of machine operation which include;
Normal Frequencies: Vibration peaks that appear in the spectrum as a normal condition of the
function of the machine such as running speed, fan blade pass or pump vane pass among others.
Defect (fault) Frequencies: Calculated frequencies that appear in the spectrum only when there is
a problem with the component such as a bearing defect frequency.
Fundamental Frequency: It is also called the first harmonic. It is a pre-calculated frequency, such
as turning speed of the machine or a frequency that is equal to a defect in the machine.
Dominant Peak: A peak in the spectrum with the highest amplitude value or a sharp maximum.
Harmonics: Spectral peaks that occur at integer (whole number) multiples of any frequency.
Sub-harmonic: Frequency that is integral sub-multiple of a fundamental frequency.
Sidebands: A pattern of equally spaced peaks above and below a centre calculated frequency.
Sidebands are caused by modulation of a carrier frequency.
Carrier (centre) Frequency: A vibration peak that is either changing in amplitude or frequency.
The carrier frequency is the “centre” frequency that appears between the sideband peaks.
Figure 5 below shows time domain waveform and spectrum of a roller bearing vibration signal.
The interpretation regarding the vibration of the component is based on the spectrum obtained
especially on the amplitude of vibration. If the amplitude of the vibration obtained is higher than
the allowable limits, then it implies that there is an excess vibration which can damage the
components and other attachments. This will then be used to inform the maintenance actions
which can be implemented to correct the problem and reduce the vibrations.

Figure 5: Time domain waveform and spectrum of a roller bearing vibration signal (Luo, S.,
Cheng, J. and Fu, 2015)
Figure 6 below shows the classification of damage in a machine based on vibration monitoring
and analysis. The level of vibration determined by the amplitude was used to classify the damage
on the machine and therefore informs the maintenance activities which can be taken to avoid
extra damage.

Figure 6: Damage classified based on the vibration analysis (Das, S., Saha, P. and Patro, 2016)

Process and maintenance engineers strive to achieve 100% uptime of the plant, avoiding any
major malfunctions to keep equipment running properly 24/7 and nearly 365 days a year.
Effective equipment maintenance helps to achieve this goal. However, the cost and uptime trade-
offs of preventive and reactive maintenance (i.e., repair) must be balanced. This can be achieved
via condition monitoring.
In a reactive maintenance scheme (also called run to failure), a machine or a process is allowed
to run until it fails, and then efforts are made to bring it back to its optimal or near-optimal state
to restore service. Although the initial cost is less than other maintenance methods, the repair
costs are often high, owing to lost process output and equipment damage, as well as the risk of
hazards to personnel and other equipment.
Preventive maintenance, or time-based maintenance, implements a periodic, custom-designed
maintenance schedule aimed at maintaining the machine or its components at the optimal
condition to reduce the likelihood of failure. Low rates of failure keep repair costs low, but
variable factors such as operating condition and component behavior make determining
appropriate maintenance intervals challenging
Condition monitoring is similar to preventive maintenance. It involves regular scheduled
maintenance to uphold machine condition, but with an additional feature — machine condition is
evaluated to determine the instant maintenance is needed, without which the equipment cannot
be operated without damage or failure. Equipment health is assessed by monitoring dynamic
operating characteristics, such as vibration signature and temperature, to create a profile of
variations and trends that can be used to predict the need for maintenance. Condition monitoring
significantly reduces spare parts and labor costs, and it increase machine availability —
effectively reducing net maintenance costs (Figure 1).

Figure 1. Condition monitoring strikes a balance between the cost of performing preventive
maintenance and the cost to repair or replace failed equipment under a reactive maintenance

program.

1. Select condition indicators. The success of a CBM program relies heavily on selection of the
proper condition indicators, which provide the insight necessary to evaluate equipment or
process function and failure modes at all operating conditions. Comprehensive knowledge of the
functioning and operating condition of the equipment to be monitored, as well as its failure
dependencies, is required for successful parameter selection. For example, temperature can be
used to calculate heat-transfer coefficients to indicate possible tube fouling, ultrasonic testing can
reveal corrosion degradation of pipelines and leakages, vibration analysis can indicate structural
problems, and thermographic scans can identify unusual temperature conditions and failing
insulation. The parameter(s) selected are the basis for selecting the condition monitoring
techniques to be used.
2. Acquire and analyze data. Depending on the parameter selected, specialized sensors collect
data that represent the actual physical state of the equipment. Sensors are typically evaluated and
selected based on sensitivity, precision, cost, and invasiveness. To the extent possible, sensors
should support online measurement of the parameter. Amplification and preprocessing of the
output from sensors is also important and is usually integrated in the data acquisition setup.
Once the data are collected, data analysis methods aim to detect incipient faults to provide an
indication that a comprehensive analysis is needed. The data analysis techniques, in principle,
identify any deviation from normal operating characteristics or causal relationships between
events and system failure.
3. Decide on maintenance requirements. Abnormal signals are analyzed to generate
recommendations that clearly indicate machine health and maintenance requirements. This
decision-making process involves diagnostics and prognostics. Diagnostics focus mainly on the
current state of the machine, determining whether the system is at immediate risk of a failure
and, if so, the location and the nature of the impending failure. Prognostics aim to predict faults
before they occur, estimating the time to failure given the equipment condition and past
operation profile. An engineer aided by computer-generated models or models that use artificial
intelligence (sometimes a combination of those) make the final maintenance decision.
Condition monitoring techniques. The condition indicators should characterize the function of
the entire piece of equipment or a particular component, such as the temperature of a particular
part, quality of the oil or lubricant, or vibration pattern of rotating parts. Based on the parameters
to be monitored, one or more relevant sensing and analysis technique is chosen.

Vibration sensors:
a. Displacement sensors

Used to measure shaft motion and internal clearances.

Used in low-frequency (1 to 100 Hz) measurement and measure low- amplitude
displacement typically found in sleeve-bearing machine designs.
b. Velocity sensors
Used for low to medium frequency measurements (1 to 1000 Hz)
Useful for vibration monitoring and balancing operations on rotating machinery.
c. Accelerometers
Are the preferred motion sensors for most vibration monitoring.
They measure low to very high frequencies
Are available in a variety of general purpose and application-specific designs.
Factors to consider when selecting vibration sensors:
Measuring range, frequency range, accuracy, transverse sensitivity and ambient conditions.
Figure: Vibration monitoring of a gearbox
Temperature measurement
Temperature is a characteristic feature of any electrical or mechanical process. Temperature-
based condition monitoring helps to detect potential faults that produce thermal changes.
Abnormal thermal variations can reveal problems such as excessive mechanical friction
(e.g., due to worn bearings or inadequate lubrication), reduced heat transfer (e.g., heat exchanger
fouling), or poor electrical connections (e.g., loose, corroded, or oxidized connections or short
circuits).
Two types of sensing technology are used to measure temperature profiles: contact and
noncontact. Contact methods of temperature measurement generally include thermocouples,
thermistors, bimetallic thermometers, and resistance temperature detectors (RTDs) (Table 1).

Table 1. Each contact temperature sensor has advantages and disadvantages,

which should be evaluated to determine the suitability of the sensor for an
intended application.

Sensor Advantages Disadvantages

 Rugged
Susceptible to noise
Immune to mechanical
Thermocouple Low and nonlinear
stress
output
Self-excited

Requires an auxiliary
High sensitivity
power source
Thermistor
Quick response
Cannot be interchanged

Low cost
Error is introduced over
Bimetallic Thermometer Robust
time
Good accuracy

Suitable for smaller

temperature ranges
Expensive
Resistance Temperature
Detector (RTD) Linearity
Prone to mechanical stress
Stability

Most contact measurement devices require minimal skill on the part of technicians, making them
popular in industrial applications. However, noncontact measurement using infrared sensors has
become increasingly popular.
Infrared thermography. Thermal cameras can capture infrared radiation emitted by an object to
locate an abnormal heat pattern or thermal anomaly. Planck’s Law and Stefan-Boltzmann’s Law
are the underlying principles of infrared thermography. All objects with a temperature above 0 K
emit electromagnetic radiation in the infrared region of the electromagnetic spectrum, and the
intensity of the radiation emitted is a function of the temperature of the object. The radiation is
absorbed by a thermal camera, which converts the infrared signature into a colored image or
video that can be analyzed to identify temperature variations. Comparing images of the same
object taken at different times is an effective way to spot abnormalities.
Infrared thermography is generally classified as either passive or active thermography. In passive
thermography, the temperature gradients are naturally present in the materials and structures
under test. However, in some cases, the thermal gradient is not prominent, so deeper and smaller
defects may not be visible using passive thermography. Active thermography overcomes this
challenge by using an external stimulus to induce thermal contrasts. Passive thermography is
used mainly for condition monitoring of electrical and mechanical equipment, whereas active
thermography is used in gas leak detection, agriculture, biology, and thermal efficiency surveys
of buildings.
Infrared thermography can be used to locate major electrical faults. For example, under normal
conditions, the temperatures (which correspond to phase) of all three electrical inputs to a three-
phase junction box are typically the same. If a fault exists in one or more phases, it will likely be
accompanied by an abnormal rise in temperature at the faulty connection (Figure 2).

▲Figure 2. The high temperature at a connection in this three-phase junction box indicates an
electrical fault.

Faults in the bearings of rotating machinery can lead to huge economic losses, so they require
timely detection. Figure 3 compares thermal images of a motor’s nondrive end (NDE) bearing
cap (Figure 3a). Figure 3b shows the normal temperature profile of the healthy motor. The white
spot in the center of Figure 3c is evidence of an abnormally high temperature.
▲Figure 3. An external temperature rise at the nondrive end (NDE) bearing cap of a motor (a) can
be detected by comparing normal (b) and abnormal (c) thermographic images. Temperature changes
could indicate a problem.

Regularly monitoring the temperature of critical components and looking for sudden abnormal
temperature increases can be useful in detecting the onset of potential faults.
Infrared thermography offers several advantages. It provides noncontact online, continuous, and
real-time monitoring, and it can be applied to a wide variety of applications (e.g., electrical
faults, mechanical inaccuracies, insulation problems, leaks). However, infrared cameras are
considerably more costly than some other measurement devices. (Editor’s Note: Read the
January 2017 CEP article “Use Thermography to Expose What’s Hidden,” pp. 21–27, for more
examples of thermographic troubleshooting.)

A condition-monitoring framework
The framework for a condition-based monitoring (CBM) program includes three major steps:
select the appropriate condition indicators
acquire and analyze data using appropriate sensing techniques
decide whether maintenance is needed and, if so, what type.
1. Select condition indicators. The success of a CBM program relies heavily on selection of the
proper condition indicators, which provide the insight necessary to evaluate equipment or
process function and failure modes at all operating conditions. Comprehensive knowledge of the
functioning and operating condition of the equipment to be monitored, as well as its failure
dependencies, is required for successful parameter selection. For example, temperature can be
used to calculate heat-transfer coefficients to indicate possible tube fouling, ultrasonic testing can
reveal corrosion degradation of pipelines and leakages, vibration analysis can indicate structural
problems, and thermographic scans can identify unusual temperature conditions and failing
insulation. The parameter(s) selected are the basis for selecting the condition monitoring
techniques to be used.
2. Acquire and analyze data. Depending on the parameter selected, specialized sensors collect
data that represent the actual physical state of the equipment. Sensors are typically evaluated and
selected based on sensitivity, precision, cost, and invasiveness. To the extent possible, sensors
should support online measurement of the parameter. Amplification and preprocessing of the
output from sensors is also important and is usually integrated in the data acquisition setup.
Once the data are collected, data analysis methods aim to detect incipient faults to provide an
indication that a comprehensive analysis is needed. The data analysis techniques, in principle,
identify any deviation from normal operating characteristics or causal relationships between
events and system failure.
3. Decide on maintenance requirements. Abnormal signals are analyzed to generate
recommendations that clearly indicate machine health and maintenance requirements. This
decision-making process involves diagnostics and prognostics. Diagnostics focus mainly on the
current state of the machine, determining whether the system is at immediate risk of a failure
and, if so, the location and the nature of the impending failure. Prognostics aim to predict faults
before they occur, estimating the time to failure given the equipment condition and past
operation profile. An engineer aided by computer-generated models or models that use artificial
intelligence (sometimes a combination of those) make the final maintenance decision.
Condition monitoring techniques. The condition indicators should characterize the function of
the entire piece of equipment or a particular component, such as the temperature of a particular
part, quality of the oil or lubricant, or vibration pattern of rotating parts. Based on the parameters
to be monitored, one or more relevant sensing and analysis technique is chosen.
Elements of CBM

Dynamic monitoring
Dynamic monitoring schemes measure and analyze the energy released from mechanical
processes in the form of waves, such as vibrations, pulses, and acoustic effects. The frequencies
of these vibrations are directly related to the geometry and condition of equipment or processes.
Any changes in dynamic characteristics can be precursors of imbalance, misalignment,
resonance, structural failures, bearing faults, or other developing problems.
Vibration signature analysis. Every operating machine produces a characteristic signature of
vibrations, which should not change over time. However, when a structural or functional
anomaly progresses, the dynamic characteristics of the machine change and are reflected in its
vibration signature. The nature of a developing fault has unique vibration characteristics, which
can be compared with the vibration signature of the equipment under normal operating
conditions. Various signal analysis techniques can be used to determine the category or type of
fault.
Vibration of a machine can be represented in a time domain (top) or in a frequency domain
(bottom). Vibration signals of rotary machine systems, such as machine tools, wind turbines, and
electric motors, can be broadly classified as stationary or nonstationary. Stationary signals are
characterized by time-invariant statistical properties such as periodic vibrations (caused by, for
example, a worn bearing), and they can be analyzed by simple techniques such as a Fourier
transform. Nonstationary signals are transient, with durations generally shorter than the
observation interval. Nonstationary signals are generated by faults such as sudden breakage of a
drill bit, flaking of the raceway of a rolling bearing, or the growth of a crack. Analysis of
nonstationary vibration signals is conducted using time-frequency techniques, such as a short-
time Fourier transform (STFT), wavelength transform, and Hilbert-Huang transform (HHT).
PHM Costing
Effect of Maintenance Strategies
Effect on:
Availability
Number of maintenance actions
Useful lifetime
RUL
Potential Cost Avoidance Associated with PHM
Failures avoided
Minimizing the cost of unscheduled maintenance
Increasing availability
Reducing risk of loss of system
Increased human safety
Minimizing loss of remaining life
Minimizing the amount of remaining life thrown away by scheduled maintenance actions
Logistics (reduction in logistics footprint)
Better spares management (quantity, refreshment, locations)
Lead time reduction
Better use of (control over) inventory
Minimization of investment in external test equip
Optimization of resource usage
Repair
Better diagnosis and fault isolation (decreased inspection time, decreased trouble shooting time)
– Reduction in collateral damage during repair
– Reduction in post-repair testing
Reduction in redundancy (long term)
– Can redundancy be decreased for selected sub-systems?
Reduced waste stream
– Less to end-of-life (dispose of) – disposal avoidance
– Reduction in take-back cost
Reduction in no-fault-founds
Eases design and qualification of future systems Reduced liability
Warranty claim verification (resulting in warranty reserve fund reduction)

Time to Failure Distributions

TTF distribution is a cumulative distribution function which describes the probability of failure
up to and including time t, where t is the failure time.
If the test was run until all instances of the product failed, then the area under the curve is 1 and
we have a PDF of failures
TTF distribution represents variations in manufacturing and materials
Cost per unit time for scheduled vs unscheduled maintenance
– Scheduled maintenance →result of diagnostics and prognostics

Evaluating Return on Investment

What is ROI?

Why evaluate the ROI?

To build a business case for implementation
To perform cost/benefit analysis on different prognostic approaches
Evaluate when PHM may not be warranted
Interpreting ROI:
0 = breakeven (no cost impact)
> 0 there is a direct cost benefit
< 0 there is no direct cost benefit
Formulating an ROI for PHM
ROI relative to unscheduled maintenance gives

Investment cost

Combine Adjustments and Maintenance for Harvest

Combine Calibration for Accurate Yield Data
Introduction
A combine is a complex machine that gathers, threshes, and cleans grain. Combines need to be
properly equipped and adjusted for the crop to be harvested. Adjustments and maintenance are
important to minimize harvest losses and deliver a high-quality grain crop. The goal of proper
adjustments is to achieve a smooth, even flow of crop material moving through the combine.
Start with the manufacturer’s recommended settings and make adjustments according to the
condition of the crop.
Safety and Maintenance
Always work safely around combine equipment and follow safety instructions in the operator’s
manual. Before harvest, follow the maintenance checklist provided by the manufacturer and
inspect for any worn parts. Check the roller chains on a corn head or the knives on a cutterbar
head for soybeans and small grains. In the threshing and separation area, check the rotor and
concave for wear or damage. Check the condition of the sieves in the cleaning shoe area. Check
for sharp edges on all augers which can lead to grain damage. Replace any worn parts as
necessary for an efficient harvest.
Header Adjustments
The header is the first contact point with the crop and can be the largest source of grain loss. The
combine header cuts and gathers the crop. On a corn header, gathering chains pull stalks back
into the header, stripping rolls pull the stalks down, and deck plates pop ears off the stalks.
Stripping roll spacing should be set according to stalk thickness. Deck plates need to be properly
adjusted to minimize ear and kernel loss. Set deck plates as wide as possible without losing ears
or shelling corn off the ear. Gathering chains pull ears into a cross auger that delivers them to the
center of the head where they enter the feeder house. Auger clearances must be properly set to
work efficiently. When harvesting, it is important to match the feeder house and corn head
gathering speed to the combine drive speed. If the gathering speed is too slow, corn stalks
entering the header will be pushed forward and ear loss can occur. If the gathering speed is too
fast, ears can be damaged with kernel loss as they impact deck plates. Speed settings are
correctly matched when stalks are pulled straight down after entering the header and ears are
gently snapped off without damage. Watch for changes in crop and field conditions while
harvesting and make adjustments accordingly.
Feederhouse Adjustments
The feederhouse is where the grain first enters the combine. Proper adjustments and settings
need to be made according to the crop. It is important to set the proper height position of the feed
drum to prevent grain damage. The feed accelerator should also be set at low speed so that whole
ears are moved into the combine to start the threshing process. If the accelerator speed is set too
fast, corn cobs can break apart and kernels can be lost before the threshing process begins.
Threshing Adjustments
Adjustments to threshing and separating parts of the combine are important for an efficient
harvest. Threshing requires a balance between rotor speed and concave clearance. The cylinder
or rotor speed is the leading cause of grain damage by the combine. Use the lowest possible rotor
speed that will shell the grain. When setting the rotor speed and concave clearance, begin with
factory recommended settings as a starting point and fine-tune for the crop and field conditions.
A properly adjusted rotor speed and concave clearance will detach most of the grain from the
cob. If the rotor speed is too fast, corn ears will break apart and kernel damage can occur
resulting in cracked grains and fines entering the grain tank. If the concave clearance is too wide,
complete threshing will not occur and similar results as excessive rotor speed will occur with lots
of split cobs showing up in the tank. For good threshing and separation, follow these guidelines:
Keep the rotor chamber full to minimize harsh grain contact,
Keep rotor speeds as low as possible for proper threshing to occur,
Close the concave spacing to increase threshing capability, and
Only increase the rotor speed as a last resort.
Grain Cleaning Adjustments
After threshing, the grain is separated from the non-grain crop material by the chaffer and shoe
sieves and the cleaning fan. The chaffer is the upper sieve that allows all grain and un-threshed
cobs or pods through to the shoe sieves which allows only grain to pass. Un-threshed crop
portions go into tailings and return to the rotor cage for re-threshing. Sieve and fan settings are
critical to deliver a clean, high-quality end product and should be set to specifications. If the fan
speed is too low, you will see pieces of stalks and cobs in the grain tank. If the fan speed is too
high, you can lose grain out the back of the combine. Sieve adjustments should work in tandem
with fan speed for best results.
Summary
Modern combines are being engineered to automate and make the adjustments easier to do, but it
still does not alleviate the need for operators to make adjustments according to field and crop
conditions. It is generally recommended to make only one adjustment at a time before making
another and making them in small increments. Further adjustments are generally necessary while
harvesting when field and crop conditions change, such as grain moisture content. Monitor losses
behind the combine and grain quality in the grain tank throughout the day while harvesting.
Always refer to the owner’s manual for complete information for combine maintenance and
adjustments.

Today’s combines can harvest up to 30 acres per hour and only require one person to operate
them. They are highly sophisticated machines fitted with advanced electronics that enhance
performance, precision and operator comfort. Like modern tractors, this sort of technology is
crucial for boosting global food production to meet ever-growing demand while also promoting
long-term environmental sustainability.
10 vital electronic systems in modern combine harvesters
1. Engine control unit (ECU)
The ECU is the brain of the combine harvester’s engine, managing and optimising performance
by monitoring and controlling fuel injection, air intake and other critical parameters. This
sophisticated component enhances fuel efficiency, reduces emissions and ensures optimal engine
operation under various conditions – this makes it an indispensable part of modern harvesting
machinery.
2. Hydraulic control unit (HCU)
The HCU is responsible for managing the harvester’s hydraulic systems, controlling various
components and attachments. It oversees lift, tilt and other hydraulic functions, greatly
enhancing the versatility and efficiency of the harvester. This allows for the seamless operation
of multiple harvesting functions simultaneously.
3. GPS and navigation systems
These systems enable precision farming by providing accurate guidance and mapping for
harvesting operations. They support auto-steering capabilities, reduce overlap and ensure precise
harvesting patterns. This technology not only increases efficiency but also minimises fuel
consumption and crop damage.

4. Telematics and connectivity modules

These modules facilitate communication between the harvester and external systems, enabling
remote monitoring and diagnostics. They allow for real-time monitoring, predictive maintenance
and over-the-air firmware updates, keeping the harvester in optimal condition and minimising
downtime.
5. Grain loss monitors
Even with the best equipment, there are always going to be small losses of crops during
harvesting unfortunately. Electronic sensors detect and quantify grain loss during the harvesting
process, allowing operators to adjust settings in real-time. This technology helps to minimise
losses and improve yield, ensuring that farmers get the most out of their crops.
6. Moisture sensors
These sensors measure the moisture content of the harvested grain, ensuring that crops are
harvested at the optimal moisture level. This prevents spoilage and maximises storage quality,
which are crucial for maintaining the value of the harvest.
7. Yield monitoring sensors
By measuring the flow and quantity of harvested crops, these sensors provide real-time yield
data. This information helps farmers to monitor crop performance, optimise harvest strategies
and make informed decisions for future planting. This sort of ‘bigger picture’ data is vital for
understanding what is happening on a farm and for contributing to long-term productivity.
8. Electronic power steering (EPS)
EPS provides electronic assistance to the steering mechanism, significantly improving the
manoeuvrability of these large machines. By reducing driver fatigue and enhancing steering
precision, EPS allows operators to safely work longer hours with greater accuracy, which
improves overall harvesting efficiency.
9. Brake control module
This module manages the harvester’s electronic braking systems, including brake force
distribution. It ensures effective and balanced braking, enhancing safety during harvesting
operations, especially when working on uneven terrain or in challenging weather conditions.
10. Traction control system
The traction control system optimises traction in varying field conditions by preventing wheel
slippage. This improves stability and performance in difficult terrains, allowing harvesters to
operate efficiently in a wide range of environments and soil conditions.
A supporting role
On top of this, there are also several other components that contribute to greater efficiency,
higher yields and increased safety:
Electronic grain tank management: This system automates the management of the grain tank,
including filling and unloading processes. By increasing efficiency and reducing the risk of
overfilling or spillage, electronic grain tank management ensures smoother operations and
minimises grain loss.
Battery management systems: These systems monitor and manage the harvester’s electrical
system, ensuring efficient use of electrical power and extending battery life. By optimising
power distribution and preventing overcharging or deep discharging, battery management
systems contribute to the overall reliability and longevity of the harvester’s electrical
components. This results in reduced maintenance costs and improved operational continuity.
Throttle-by-wire system: This electronic system controls the engine throttle for improved
responsiveness. By enhancing fuel efficiency and providing smoother engine performance,
throttle-by-wire technology contributes to more precise control over the harvester’s speed and
power output. The outcome is better overall harvesting performance and reduced fuel
consumption.
Remote control systems: Remote control capabilities allow for operation and monitoring of the
harvester from a distance. This technology supports autonomous or semi-autonomous features,
increasing efficiency and also reducing labour costs. Remote control systems can be particularly
beneficial for large-scale operations or in situations where the presence of an operator might be
challenging or unsafe.
Figure 1. A combine in a field ready to harvest corn.

Camera systems: Modern harvesters are equipped with rear-view and side-view camera systems
for enhanced visibility. These cameras increase safety and precision during operations, allowing
operators to maneuver the large machine with greater confidence. They are particularly useful
when reversing or operating in tight spaces, reducing the risk of accidents and improving overall
efficiency.
Climate control systems: Modern combine harvesters feature sophisticated climate control
systems that manage air conditioning, heating and ventilation within the cab. These systems
significantly enhance operator comfort, allowing for extended working hours in any weather
conditions. By maintaining an optimal working environment, they contribute to increased
productivity and reduced operator fatigue.
Instrument cluster and display: This provides the driver with essential information such as
engine parameters, fuel levels and diagnostic alerts. By centralising this critical data, it improves
operational awareness and leads to better decision-making. Farmers can quickly assess the
harvester’s status and performance, enabling them to make timely adjustments and maintain
optimal efficiency.
Data is key
The integration of advanced electronic components in modern combine harvesters has
revolutionised the agricultural industry. From ECUs and GPS navigation, to yield monitoring
sensors and remote control systems, each component plays a crucial role in optimising the
harvesting process. These technologies not only improve the quantity and quality of harvested
crops but also contribute to sustainable farming practices by reducing waste and minimising
environmental impact.
The importance of gathering and interpreting farming data is more important now than ever
before. The data collected by these electronic systems provides valuable insights that feed into
long-term agricultural strategies, potentially transforming how we approach crop management
and farm planning.
As technology continues to evolve, we can expect ever-more innovative electronic components
to be integrated into combine harvesters. Advances in areas such as AI, machine learning and
Internet of Things (IoT) technologies are likely to further transform the way we harvest crops.
These developments will bring about even greater levels of automation, precision and efficiency
in agricultural operations. This will help to pave the way for a new era of smart farming, which
can help to address global food security challenges while also promoting sustainable agriculture.