Q 539. There are 3 major types of maintenance methods - Corrective, Preventive and Predictive. While we all understand Corrective and Preventive maintenance, what is the concept of Predictive maintenance? How is it different from other two and what are the tools used for predictive maintenance?

 

Note for website visitors - Two questions are asked every week on this platform. One on Tuesday and the other on Friday.

• All questions so far can be seen here - https://www.benchmarksixsigma.com/forum/lean-six-sigma-business-excellence-questions/
• Please visit the forum home page at https://www.benchmarksixsigma.com/forum/ to respond to the latest question open till the next Tuesday/ Friday evening 5 PM as per Indian Standard Time. Questions launched on Tuesdays are open till Friday and questions launched on Friday are open till Tuesday. 
• When you respond to this question, your answer will not be visible till it is reviewed. Only non-plagiarised (plagiarism below 5-10%) responses will be approved. If you have doubts about plagiarism, please check your answer with a plagiarism checker tool like https://smallseotools.com/plagiarism-checker/ before submitting. 
• The best answer is always shown at the top among responses and the author finds honorable mention in our Business Excellence dictionary at https://www.benchmarksixsigma.com/forum/business-excellence-dictionary-glossary/ along with the related term

Predictive Maintenance

 

Predictive maintenance is a type of maintenance strategy that uses data and analytics to predict when a piece of equipment is likely to fail. The goal of predictive maintenance is to identify potential failures before they occur, so that maintenance can be performed proactively to minimize downtime and reduce the risk of unexpected equipment failures.

 

Predictive maintenance is different from corrective maintenance in that corrective maintenance is performed after a problem has occurred, while predictive maintenance is performed before the problem occurs. Preventive maintenance, on the other hand, is performed at regularly scheduled intervals regardless of whether a problem has occurred or not.

 

Corrective Maintenance	Preventive Maintenance	Predictive Maintenance
Done after the failure or incident – Reactive	Done prior to avoid failure or incident - proactive	Done to forecast potential issues so that they can be avoided - proactive
Does not involve any planning, cost effective in short term	Lot more complex planning. Scheduled basis system run time. Expensive to begin with	Scheduled basis need using real time condition of system. Relative more complex to implement. Can be very expensive to begin with depending on sensors or tools reqd.
Process is simple as its done only when needed. At other times the maintenance team can focus on other activities	Labour intensive and maintenance team always has some assets to work on. Maintenance carried out whether reqd. or not	Performed only as needed so more efficient
Very expensive on long term as the failure or incident would cause damage to system. Asset lifecycle is reduced	Expensive but avoids damage to system. Asset life cycle is improved	Over long term the maintenance cost will be minimum. More efficient in utilizing asset life before replacing
Has risk of injury for employees	Better for safety of employees	Better for safety of employees
Downtime of assets and therefore impact on productivity is high	Downtime of assets is low	Downtime of assets is minimum
Since it only focuses on issue and not the cause, chances of repeat failure are high	Chances of failure is low	Chances of failure is very low
Can be suitable when cost of part is significantly cheaper than cost of predictive or preventive maintenance	Suitable when cost of failure or breakdown or repair is very high	Benefit of preventive maintenance but more efficient and cost effective

 

 

There are several tools used for predictive maintenance, including:

 

·         Vibration analysis: This involves measuring the vibration of machinery to identify issues such as misalignment, imbalance, or bearing wear.

·         Thermography: This involves using infrared cameras to detect changes in temperature that can indicate potential problems with electrical equipment, mechanical equipment, or other components.

·         Oil analysis: This involves analyzing the oil used in machinery to identify contaminants and wear metals that can indicate potential problems with the machinery.

·         Machine learning algorithms: This involves using data from various sensors and sources to build predictive models that can help predict equipment failures.

·         Condition monitoring software: This software collects and analyzes data from various sensors and sources to provide real-time information on the health of equipment, which can be used to identify potential problems before they occur.

·         Ultrasonic testing: This involves using high-frequency sound waves to identify potential problems with equipment such as cracks or corrosion.

·         Electrical signature analysis: This involves analyzing the electrical signals produced by machinery to identify potential problems with the equipment's electrical components.

·         Wear debris analysis: This involves analyzing particles generated by the wear of machinery to identify potential problems with the equipment's components.

These are some of the most commonly used tools for predictive maintenance, but there are many other tools and techniques that can be used depending on the specific needs of the equipment and the organization.

 

Predictive maintenance (PdM) / Condition-based maintenance (CBM) While Preventative and corrective maintenance emphasizes on Age of the asset and frequently scheduled maintenance, the constraint is a focus on only repairing assets at the last stage of degradation.

Predictive maintenance lets you monitor equipment health to avoid upcoming failures during the process. It uses predictive algorithms with data from equipment sensors to evaluate when the equipment will fail. It also identifies the root cause of problems in your difficult machinery and helps you identify which portions need to be repaired or replaced. This way, you can reduce downtime and make the best use of the equipment’s lifetime.

 

Predictive maintenance software uses data science and predictive analytics to estimate when a portion of equipment might fail so that corrective maintenance can be planned before the point of failure. The goal is to schedule maintenance at the most appropriate and most cost-efficient moment, letting the equipment’s lifespan be improved to its fullest, but before the equipment has been compromised.

The predictive maintenance solutions typically consist of data procurement and storage, data transformation, condition monitoring, asset health evaluation, prognostics, a decision support system, and a human interface layer.

 

Difference between Preventive vs Predictive Maintenance

The difference between preventive maintenance and predictive maintenance lies in the data being analyzed. While a predictive maintenance specialist relies on monitoring and analyzing data from the actual, current condition of the equipment in operation, preventive maintenance relies on historical data, averages, and life expectancy statistics to predict when maintenance actions will be required.

Corrective maintenance refers to the actual repair or replacement of equipment that has failed, broken, or worn down.  Preventive, predictive, and corrective maintenance is particularly critical for safe operations in the Oil and Gas, Manufacturing, Telecommunications, and Railway industries.

 

How to Implement Predictive Maintenance?

Manufacturers are increasingly using the Internet of Things (IoT) predictive maintenance and AI predictive maintenance services to implement automatic predictive maintenance in their operations and equipment. AI in predictive maintenance can tailor maintenance routines to the requirements of each individual piece of equipment and can be trained to visually identify flaws and patterns in equipment.

 

Why Predictive Maintenance is Important?

Predictive maintenance insights are an exceptionally valuable asset in refining the overall maintenance and reliability of an operation.

 

Benefits include:

 

•                 minimalize the number of sudden breakdowns

•              Utilize asset uptime and improve asset reliability

•              reduce operating costs by carrying out maintenance only when needed

•              maximize production hours

•              Improve safety

•              Streamline maintenance costs by reducing equipment, inventory costs, and labor.

 

Top 5 Predictive Maintenance Software

•                  Limble CMMS.

•              Fracttal One.

•              Fiix.

•              SAP Predictive Asset Insights.

•              Brightly Asset Essentials.

Predictive maintenance is a type of maintenance strategy that aims to predict when equipment or machinery is likely to fail so that maintenance can be performed proactively, before the failure occurs. The goal of predictive maintenance is to minimize equipment downtime and reduce the cost of maintenance by fixing problems before they become serious.

 

Predictive maintenance is different from corrective maintenance, which only takes place after equipment has failed, and preventive maintenance, which is performed on a regular schedule regardless of the condition of the equipment.

 

Predictive maintenance uses various tools and technologies to monitor equipment performance and predict when maintenance is needed. Some of the tools used for predictive maintenance include:

 

1. Condition monitoring equipment: This includes sensors and monitoring devices that collect data on the performance and condition of equipment. This data can be used to identify potential problems before they become serious.

 

2. Machine learning algorithms: Predictive maintenance often utilizes machine learning algorithms to analyze data collected from condition monitoring equipment. These algorithms can identify patterns in the data and make predictions about when maintenance is needed.

 

3. Predictive maintenance software: This software can help automate the process of predictive maintenance by collecting and analyzing data, making predictions, and generating maintenance schedules.

 

4. Internet of Things (IoT) devices: IoT devices can be used to monitor equipment performance and transmit data in real-time to a centralized system for analysis.

 

By using these tools and technologies, predictive maintenance can help organizations reduce downtime, improve equipment reliability, and minimize maintenance costs

Whether a corrective, preventive, or predictive maintenance plan is most appropriate for a given product or asset must be determined by maintenance works for any type of asset. Selecting the proper maintenance activity necessitates understanding of each strategy and how it could affect your resources and work schedule because each approach to maintenance has advantages and disadvantages of its own.

 

Equipment breakdowns can be expensive, thus the most logical course of action may seem to be to implement a predictive maintenance policy that anticipates failures and allows for the implementation of corrective action in advance. However, some non-essential equipment might wear out without actually being an issue and can often be quickly and affordably fixed. A maintenance crew may be inclined to acknowledge a corrective maintenance technique under such circumstances. Run to breakdown may also be used when remedial action is impractical. For instance, some spacecraft and satellites may be built with no maintenance assumptions and allowed to operate until they fail (and then abandoned).

 

Preventive maintenance is a third sort of technique that is frequently employed by business. This strategy aims to stop a failure before it happens, similar to predictive maintenance, but there is a small variation between the two, which we go into more depth about below.

 

Management software, such as CMMS Software (Computerized Maintenance Management System), which may help monitor resources in actual time, schedule maintenance, and track work orders, can help organize a condition-based maintenance strategy.

 

Predictive maintenance

 

Because it uses a more thorough method to determine when maintenance is required, predictive maintenance varies slightly from preventive maintenance. Predictive maintenance employs analytics and ongoing data monitoring to identify whether mechanical failure is probable to appear rather than using a specified time or usage pattern to decide when repair is necessary.

 

This kind of monitoring enables maintenance to be carried out as needed to address a particular issue and stop an asset from malfunctioning. Predictive maintenance is frequently more cost-effective than preventative maintenance since it only alerts the user when a malfunction is about to occur.

 

Benefits of Predictive maintenance

 

Lower spending

 

This method eliminates the excessive maintenance costs related to preventative maintenance because maintenance is only carried out when necessary. Since faults are fixed before total equipment failure occurs, maintenance chores are frequently less expensive than with reactive maintenance. By being able to schedule repairs and downtime for equipment, this not only cuts down on the cost of the time spent performing maintenance, but it can also cut down on other costs, including overtime.

 

Higher reliable asset performance

 

If your equipment is properly maintained, it will be more dependable and perform better. By ensuring that your machinery is operating properly, a predictive maintenance solution can lengthen the life of your equipment. Countless other benefits of preventative maintenance, including as improved safety, cost control, energy savings, and reduced disturbance to work schedules, are included in this increased reliability.

 

Reduced downtime

 

This maintenance method can significantly cut down on downtime or outages brought on by failing equipment. Smaller fixes and tune-ups help stop bigger issues from arising, keeping output and customer satisfaction high.

 

Routine Maintenance

 

You can schedule maintenance because you are in command of your equipment's condition. Instead of getting an expensive reactive approach, you can make sure the necessary specialists but also components are on-site when there is time to make repairs. You can also minimize interruptions to workflow to increase productivity and profits.

 

Product Quality Improvement

 

Machines that are not properly maintained can malfunction and produce items that are defective. By keeping an eye on your machinery, you can make sure that every component is working properly and continuously producing high-quality goods.

 

Predictive maintenance have some drawbacks as follows

 

Increasing Initial Costs

 

You will pay more up front for a preventive maintenance program than you would if you let anything wear out on its own. Regular maintenance costs money in terms of time, wages, and parts. You must determine whether this is higher than the possible cost of allowing something to fail naturally, though.

 

Requires knowledge of data interpretation

 

To evaluate the data from your predictive maintenance technology, you'll need qualified personnel. It's crucial to perform your vital data analysis appropriately in order to accurately identify when a problem is about to occur or maintenance is necessary. The equipment and the quality monitoring both require an understanding on the part of your technicians. This can necessitate hiring new employees or undergoing staff training, but these expenses might end up being cost-effective in the medium to long term.

 

Despite the inherent benefits of preventive and predictive methods, corrective maintenance may in some circumstances be the best option. Corrective maintenance is frequently the recommended course of action when the expenditure of part failure and replacement is less than the expenses of preventive or predictive maintenance.

 

Therefore, in the situation of a light bulb, for instance, remedial action is unquestionably the best choice because the expense and time required to change the bulb are not expected to have a significant impact on finances or work schedules. This alters, though, as the component is required for multiple functions, like in the case of a engine’ ships or a blade for a wind turbine.

 

Corrective maintenance can indeed be especially harmful when it doesn't actually been selected as a course of action; allowing something to fail inadvertently can be expensive in terms of time, money, and safety.

 

When the cost of waiting for something to break is too high, a part is too important, difficult to replace, or could seriously affect health and safety of employees or work schedules, preventive monitoring is used.

 

Although this method is not required for all things, it enables the maintenance of that are more significant. This will lessen the possibility of unplanned outages, increase the asset's lifespan and performance, and make it easier to identify any unanticipated causes of failure.

 

Predictive maintenance is becoming the strategy of choice for asset owners. It not only provides many of the advantages of preventive maintenance, but it also performs it in a more effective and economical way. A condition-based control method can be used to notify you only when measures are needed, minimize unnecessary maintenance fees while preserving your equipment in excellent condition. Although there are expenses associated with setup, after these have been budgeted for, they can be used. Sensors can be employed to remotely control a variety of operating factors, including oil quality, vibration, and cracking.

 

As technology evolves, monitoring and data collecting performance improves at a lower cost, this kind of maintenance is becoming increasingly accessible.

 

 

Which monitoring strategy to use will depend on the organization and the application. While taking a corrective, proactive approach with some objects is completely appropriate and even preferred, many will need this kind of routine maintenance or monitoring. It holds true for every single thing, from little machine parts to huge buildings or structures.

 

The frequency of maintenance examination is determined by preventive maintenance, which might be based on prior inspection data. Maintenance needs for certain assets or objects may also be specified by equipment manufacturers, laws, and even standards. These maintenance plans may need to be modified over time to improve their cost- and maintenance-effectiveness.

 

Motor circuit analysis, laser-shaft alignment, infrared analysis, oil analysis, Vibration analysis and ultrasonic analysis are the six key tools available for predictive maintenance.In order to provide an even more successful method, predictive maintenance is utilizing new technology, such as the artificial intelligence. Among the most popular leading Predictive Maintenance Software include Presenso, GE Predix, and Siemens MindSphere. I personally don’t believe that installing sensors for a lightbulb in your desk lamp is an effective form of maintenance strategy, but of course, everything still relies on the exact timing of the nature of the environment or piece of equipment.

Corrective maintenance occurs after the failure occurred. It is a reactive approach, leading to unscheduled downtime, loss of productivity and higher cost to cover up for the loss of productive hours.

Preventive maintenance, relatively common method of checking the health of the system on a regular basis, i.e., once a quarter or once a year, is used for early identification and correction of problems. Though it is a proactive approach, regular maintenance costs time and money; and in many instances no defects identified. If failure happens between the preventive cycle, it leads to productivity loss and requires corrective maintenance.

Predictive Maintenance (PdM) on the other hand identifies a problem before it occurs and alerts when maintenance is required.  Also called as Condition Based Maintenance, Predictive maintenance is system of monitoring key parameters for smooth operation, identifying early warning signals and prompting for intervention. This reduces the probability of unexpected failure and improves predictability and reliability of the system. Predictive maintenance software uses a large data source to predict potential failure in advance.

Continuous monitoring and use of analytics are essential components of predictive maintenance. Various tools like IoT, big data analytics, machine learning, artificial intelligence are used to accurately predict the need for maintenance. This is performed continuously or periodically, depends on the impact of failure. Since this requires a large data collection, regular monitoring, Predictive maintenance is more expensive than other types of maintenance. In addition to higher investment, it also requires a longer duration to assess and implement the solution.

 

The term Maintenance, has been used in industrial literature, sparingly, to designate improvement initiatives, targeting – functional status check, repair and replacement of tool & machinery devices, development of infrastructure and utilities. The maintenance activities may be clubbed as:

 

Corrective maintenance: activities comprising repair attempted after a breakdown is registered or in compliance to notice shared over fault-issue-limitation observed, within process and machinery. When the machinery is not in production or undergoing repair check, during a pre-defined schedule; the defect in machinery and production line are identified. These defects are then attended to and corrected, on a future date, assigned by technical maintenance team, “just in time”, before a major defect occurs in production line.

 

Preventive maintenance: activities comprise routine schedule of periodic actives, repetitive over a time scale. Operations Strategist for a manufacturing unit, may plan several inspection visits, with series of check-lists and observation parameters; to positively identify symptoms and signs of “wear n tear”, in plant and machinery.

 

Predictive maintenance: activities comprise the use of “sensor devices”, condition-monitoring equipment, to constantly maintain vigil on state of production in manufacturing unit. As such, sensor devices are routinely monitored, to gather correct functional data, related to machinery and process health.  “Internet of Things-- IOT” is used as a tool, to achieve Predictive maintenance, where, the operational data is continuously recorded; using the sensor device RFID Radio Frequency Identification Device, Bluetooth, Zigbee, attached to machine or production unit. Here, a “Bluetooth” is a short-range wireless technology standard that permits information exchange between fixed and mobile devices over short distances. The RFID technology employs active and passive tag, out of which, the active sensor sends information over a few hundreds of meters and the passive tag receives the shared information. These RFID sensor tags are integrated in machinery, which promotes predictive maintenance and control inside the process. Zigbee is a wireless protocol enabling Smart Devices such as door sensors, light bulbs, motion sensors, plugs, smart locks and sockets to communicate between themselves over personal area network.

 

The functional data thus obtained, is transferred electronically using LAN-WAN-Cloud based platform to the customer, is monitored and analysed using Artificial Intelligence algorithms, to detect data patterns symbolizing the defective or/and steady performance, of the machine or production unit.

 

A predictive maintenance program/computer software employs condition monitoring & prognostics algorithms, to analyse data measured. Condition monitoring uses data from a machine to gauge present state of performance and to detect & diagnose faults in the machine. As such, A condition monitoring algorithm, capitulates metrics from the data called condition indicators. A conditional indicator denotes performance parameter derived from the data, that groups similar system status together, and sets different status apart. Therefore, condition-monitoring algorithm arrives at fault detection or undertakes status evaluation, by comparing new data against existing parametrised defect data. Similarly, a prognostics algorithm typically estimates the machine's remaining useful life (RUL) or time-to-failure, by monitoring present state of performance of a machine. Prognostics algorithms employ modelling, machine learning to evaluate future state of condition indicators. The future value thus obtained are used to compute remaining useful life RUL, which is then extrapolated to determine, if and when maintenance should be performed. A predictive maintenance system implements prognostics and condition monitoring algorithms with other IT infrastructure, that dictate actionable insights to end users/line managers; facilitating prerogative on maintenance schedules.

 

 

  

 

Most common of the predictive maintenance (PDM) tools include:

 

Infrared analysis	Laser-shaft alignment
Motor circuit analysis	Oil analysis
Vibration analysis	Ultrasonic analysis

 

Infrared analysis (IR) tool uses varying of levels Infra-Red light emerging from object surface in one view or multiple views, over a period of time, to indicate difference in temperature; on parts of the object surface. Here temperature difference is used to indicate machine condition or machine performance. IR tools have been used to analyse temperature difference of the machine surface, ARC flash analysis to determine electrical component condition, Insulation or building conditions, Piping and plumbing conditions, Process temperatures, Solar panel conditions.

 

Laser-shaft alignment attempts to make room for proper alignment of machine components on all three-axis of support platform. This innovation utilizing laser-shaft based alignment of machine hinge to support railing, prevents extraordinary pressure resulting from misalignment of machine drive train.

 

Oil Analysis employs analysis of oils for viscosity, water, and other wear indicators. Oil Analysis is used to detect metal fatigue basis presence of metal ion in sample oil. Many-a- times, Equipment warranty standards provides justification for the use of Oil Analysis, as a Preventive Maintenance measure. More often, Oil analysis is used on high speed or critical equipment, since sample collection can be done easily and this activity culminates to lower oil consumption by reducing periodic oil changes.

 

Motor circuit analysis analyses electric signature analysis (ESA) to identify faults in electronic motors including concerns with Incoming power, Motor electrical circuitry, Motor mechanical components and Motor mechanical couplings.

 

Vibration Analysis (VA) tool is sensor to detect vibrations emerging from a machinery. An analysis of the vibration readings provides clues to identify problem signal, changes observed between previous and current data, identification of faulty machinery pieces, identification of equipment alignment concerns and generates impetus for action plan.

 

Ultrasonic analysis (UA) tools acts upon high-frequency sounds gathered using microphone and converts it into data signals. The data may be similarly used to identify problem signal, and record changes observed between previous and current data. Newer UA tools has thermometers, cameras, and spectral analysers as add-on detection equipment, that has improved performance of Ultrasonic analysis (UA) tools, many folds. UA tools have been used to generate data for Electrical inspection, Failed steam traps and steam systems, Leak detection, Optimal lubrication practices and Valve testing.

 

 

 

Interesting answers from all participants.

 

Best answer has been provided by Anupam Goswami and Nunhuck Oosman.

 

Answer from Anshul Vaidya is also a must read.