If someone was trying to convince you NOT to start a predictive maintenance program, they would probably tell you things like:. If you too are considering starting a predictive maintenance program at your facility, read through this guide and learn about all the steps you need to take to ensure successful implementation.
Well, that depends. To be fair though, implementing a PdM program is more complicated when compared to starting a preventive maintenance program. How hard this process is going to be will depend on your level of expertise, the complexity of your assets, and the tools you have at your disposal. We recognized the arguments against predictive maintenance and looked for ways to simplify your transition to PdM. Informing yourself by reading this guide is a step in the right direction and will make the whole process significantly smoother.
Although the benefits of predictive maintenance are numerous, deployment can be challenging. So, before you start, it is important to have a well-defined plan covering the desired business outcome and the project scope.
Below are step-by-step best practices for implementing a successful and sustainable PdM strategy. First, you need to determine which assets you want to include in your predictive maintenance program. This distinction is important because of two reasons:.
By checking historic machine records — maybe over a two or three year period — you can identify assets that are most vital for your business processes and would cause a substantial disruption if they unexpectedly failed. Assets that meet the above criteria are the best candidates for the PdM program. For example, by installing vibration and motor current sensors on major electric motors and placing them on a predictive maintenance program, you can avoid breakdown of sensitive production equipment on your plant floor and avoid profit losses running into thousands of dollars per hour of downtime.
Machine records present a valuable and time-saving source of actionable data to help get PdM rolling. Such data offers information about machine behavior that will, to a large extent, determine how the PdM model is designed. New equipment always comes from the manufacturer with comprehensive manuals and instructions.
This information covers important details like what kind of maintenance is required, how often it should be done, how to perform it safely, and so on. Companies that have been keeping accurate maintenance records have an added advantage here as they can quickly gather historical data for each of the selected equipment. Machine data can be extracted from hard copy maintenance records and charts, your CMMS software, and software from other departments e.
For instance, this is how an asset report looks like in Limble:. Looking at metrics like total costs, future estimated costs, and MTBF shows you how often an asset fails which might give you a clue if an asset is worth putting on a predictive maintenance plan. They already have a working knowledge of each asset and are familiar with many failure patterns. Getting their involvement and buy-in will make it easier to pinpoint the specific problems they face and how the eventual predictive model can help.
The goal of this analysis is to identify their failure modes. It is important to focus on establishing the following:. It is used to identify failure modes and estimate the risk of each mode.
FMEA is a very detailed and thorough process that involves several worksheets and calculations but in summary, it functions like this:. The result is a prioritized list that guides the implementation team to begin work on developing the failure predictions for the highest risk assets first.
The predictive maintenance program integrates different types of machine information such as performance data, maintenance history, and design data to make timely decisions about maintenance intervention. To achieve this, it requires specific technologies and real-time equipment condition data to function effectively.
This is possible through condition-based monitoring. Condition-based monitoring is a key step in the process and it works on the assumption that all machines will deteriorate and fail partially or fully at some point.
Therefore, the goal is to preempt these failures by placing various monitoring sensors on the assets. From there, the data is collected, analyzed, and used to create predictive failure algorithms which inform your maintenance actions. There are a wide variety of sensors available including but not limited to :. Image source.A recent phenomenon called Industry 4. Accordingly, plant operators strive for increased productivity, improved operational efficiency, and better safety, given the technological tools newly available to them.
Many manufacturing facilities maintain a combination of both new and old machines. The first step to make a factory smarter is enable predictive maintenance PdM capabilities.
Predictive maintenance focuses on identifying patterns in both sensor and yield data that indicate changes in equipment condition, typically wear and tear on specific equipment. With predictive maintenance capabilities, companies can determine the remaining value of assets and accurately determine when a manufacturing plant, machine, component or part is likely to fail, and thus needs to be replaced. In the long run, every piece of manufacturing equipment will reach the end of its useful life.
In the past, companies have used schedule-based maintenance also referred as planned maintenance or time based maintenance systems In many cases the approach focuses on reducing failures and end up replacing machines that could have been service for longer, because the schedule assumes worst-case lifetimes for critical pieces of manufacturing tooling.
However, this condition-based method requires observing the equipment at regular intervals. Predictive maintenance identifies condition of machinery or equipment and determines whether a specific machine is going to fail or not.
This new paradigm is enabled by connected machines that generate large amount of instrumented data and machine learning capabilities that can derive meaning from this data. Several technology trends are accelerating the present industrial revolution by enabling predictive maintenance:.
Given these technologies, machines, factories, and manufacturing processes have become intelligent, integrated, and data-driven. Predictive maintenance techniques collect data from machines such as temperature, vibration, and sound using sensors. As you collect multiple parameters from highly instrumented equipment, it becomes important to use machine learning techniques to identify failure patterns.
With machine learning you can identify patterns that indicate upcoming failure. With a deep learning-based technique, you can also identify patterns inside a pattern that indicates failure. To build a predictive maintenance solution, you should define your use case in detail by describing what you wish to predict, its business benefits, the data signals available to you, and the hypotheses you have.
AWS re:Invent, Predictive Maintenance In Your Plant Today or the Path to a Proof of Concept
The signals and failure examples you collect need to match your use case. In this scenario, your goal is to predict whether equipment will fail in a given period or not. In machine learning terms, this is referred to as a classification problem. You should check the feasibility of your use case by validating that you see a separation pattern between normal and failure instances.
In many scenarios, humans may not be able to identify such differences, but a deep learning model might be able to find a subtle or nuanced difference. In this use case, your goal is to determine the remaining life or value of assets. In machine learning terminology, this is referred to as a regression problem. To solve this problem you need to have labeled data about machine at various stages e. In this use case, you focus on identifying abnormal, unusual, or irregular patterns, as they contrast to how you expect equipment to work in a typical fashion.
Predictive maintenance techniques seek out abnormalities in data and trigger an alert as soon as the machine produces data in an abnormal range. In the above scenarios, a human designs the predictive maintenance model.
Once the system can predict whether equipment will fail or not, a human looks at the data to make a decision. But one can also use machine learning for optimization of manufacturing processes. In these use cases, the system automatically determines ideal behavior settings within a specific context to maximize performance. Most modern manufacturing facilities house complex interactions of various mechanical and electronic controls.
Modern data centers have similar challenges.
This case study describes machine learning optimizations for Data Centers. Model applications include data center simulation to evaluate a new plant configuration, energy efficiency performance, and identifying optimization opportunities.
AWS – Invent, Predictive Maintenance In Your Plant
We recommend that companies start their AI journey by implementing classification and regression models for predictive maintenance. Once they have robust pipeline and success with machine learning, they should consider building more advanced optimization models.A global leader in electrification, automation, and digitization, Siemens AG has driven innovation across industries for nearly years.
Siemens uses an array of AWS services to carry on that tradition of transformation—bringing IIOT to railways and factories, developing intelligent infrastructure for buildings and distributed energy systems, implementing AI into its cybersecurity platform, and more. Siemens built a serverless AWS solution to analyze and reduce power plant alerts. The company provides power, medical, laboratory, and manufacturing solutions.
The company is a global electrification, automation, and digitalization leader. Siemens surveys employees quarterly using AWS machine learning technologies to translate and analyze results in less than two weeks. Siemens Mobility simplified operations and democratized its data by developing and implementing an open ecosystem for rail data integration on AWS.
A leader in transport solutions for more than years, Siemens Mobility is constantly innovating its portfolio in its core areas of rolling stock, rail automation and electrification, turnkey systems, intelligent traffic systems, and related services. The solution includes a data lake based on Amazon S3 for dataset catalogs, AWS Glue for data transformation, Amazon Athena for interactive querying, and Amazon EMR for processing and analysis of unlimited amounts of data.
Siemens Smart Infrastructure used AWS to build Building Twin from Siemens, a platform that ingests and presents building-generated data, agnostic of source sensors and systems. Siemens Smart Infrastructure provides physical products and systems, cloud-based digital offerings, and value-added services.
On AWS, our AI-driven cybersecurity platform easily exceeds the strongest published benchmarks in the world. Siemens built MindSphere, an IIoT operating system that can easily connect to as many as 80 percent of worldwide industrial automation devices to accelerate customer time to value from IIoT, on an AWS-native architecture. Siemens used AWS services to bring Junelight Smart Battery—a consumer smart battery for storing solar energy—to market in just 18 months.
Companies of all sizes across all industries are transforming their businesses every day using AWS. Contact our experts and start your own AWS Cloud journey today. As part of its strategy of providing digital transformation solutions to realize value across the entire business and embrace Industry 4. Driven by automation, IoT, and cloud computing, Siemens can now solve business problems with the data that it collects, analyzes, and monitors.
Learn how and why Siemens built MindSphere on AWS, for its own global factories and for its customers, in order to achieve world-class levels of manufacturing efficiency. Siemens uses AWS Support in order to optimize its migration and lower costs. Siemens Power and Gas Case Study. Siemens Power and Gas division was looking for ways to accelerate its data-analytics and customer-digitalization projects. After choosing Amazon Redshift, it wanted to build out the solution with best-in-class software.
Siemens Power and Gas continues to use AWS Marketplace to find, try, and buy software that sets up in minutes and can shave weeks off of project timelines through the ease of implementing software and shortened procurement cycle. Siemens Mobility built a predictive-maintenance SaaS solution, improving the availability and reliability of the Siemens-enabled trains from 87 percent to 99 percent. Siemens Healthcare Diagnostics helps advance human health through innovation.
Get Started Companies of all sizes across all industries are transforming their businesses every day using AWS. Contact Sales.ST will take part in the AWS re:Invent conference during the first week of December and show how our solutions and Amazon Web Services can come together to create powerful as well as cost-effective platforms for conditioning monitoring and predictive maintenance.
We also offer a companion dashboard DSH-PREDMNT built on AWS services to facilitate the overall setup, shorten development times, and help teams go from thinking about predictive maintenance to executing a comprehensive and sensible implementation. Unlike reactive maintenance, which deals with problems as they arise, proactive maintenance improves product quality and costs. It also offers the highest equipment availability compared to aggressive maintenance strategies that take machines offline a lot more often, sometimes even if there are no signs of upcoming failures.
There are many proactive maintenance strategies, but we will focus on two of them, condition monitoring and predictive maintenance since engineers can implement both using our predictive maintenance kit, or our Discovery kit IoT node. Condition monitoring tracks the working conditions of equipment, such as the vibrations coming out of a fan. Engineers characterize a normal behavior, and the system throws an alert if sensors detect abnormal values.
Very often, the machine is still functioning, but condition monitoring helps anticipate an upcoming failure. Teams can more efficiently plan maintenance operations. Predictive maintenance goes a step further by offering a more complex data analysis system through the use of a neural network trained to more accurately forecast issues instead of just relying on thresholds.
The inference can take place on the gateway, on the sensor board, or even on the sensor itself if the latter embarks a decision tree. One of the biggest challenges that a company faces when implementing proactive maintenance strategies is finding a solution that can easily retrofit existing machines. The STEVAL-BFAV1B sensor node can monitor current equipment and process critical vibration data locally since it can output frequency and time domain data, as well as information on the temperature, pressure, and humidity.
The development board thus offers one of the most straightforward paths to a proof of concept. Similarly, if engineers have a trained neural network available, they can rapidly move into predictive maintenance thanks to STM32Cube. AI, which converts a neural network into STM32 code. Being able to run inferences on an embedded system efficiently makes its integration within a plant a lot easier.
Coming up with a proof of concept starts by understanding the potential of a cloud-based solution. AWS IoT Greengrass can simulate a cloud to keep the platform functional until the system is back online. It can even process data locally, run updates, or take some actions instead of entirely relying on the Internet.
Using an STM32MP1 gateway and AWS is thus about making more devices smarter and reliable to ensure that both condition monitoring and predictive maintenance solutions are more robust and flexible.Practical Machine Learning for Predictive Maintenance
Until recently, gateways to the cloud often required expensive and powerful machines with hypervisors or complex administration requirements. Once the system is in place, developers can write code in the cloud and deploy it remotely to increase their productivity. Engineers looking to build a test platform quickly will gravitate toward the starter package while the developers working on the final version of the software will use the distribution package to customize the operating system massively.
As engineers evaluate their proof of concept, one of their first challenges is to connect the sensors nodes and the gateway. Wired connections offer a high level of security and make snooping attacks a lot harder. Wireless technologies can make it easier to retrofit a solution onto older machines and to link assets that are far from one another.
When using IO-Link, data from up to four sensor node transceivers can route their stream through an IO-Link master to a gateway that will consolidate the information before the cloud processes it. Multiple wireless smart sensor nodes can either send their results to the gateway or immediately transmit them to the cloud. We also offer libraries to help programmers take advantage of the MCU present on their sensor node to perform a fast Fourier transform or sensor fusion computations.
It can be valuable for engineers looking to reduce the amount of data transmitted to the cloud or use the available computational power to process information quicker. Hence, we see that developers not only have the boards, firmware, and tools at their disposal to start implementing condition monitoring or predictive maintenance in their plants, they also have the flexibility to adapt it to their particular use cases.
In Application Examples. Proactive Maintenance: Understanding Condition Monitoring and Predictive Maintenance Combining ST and AWS for predictive maintenance Condition monitoring tracks the working conditions of equipment, such as the vibrations coming out of a fan.By using AWS, Hubble Connected is helping people transform their households by installing connected devices that enable owners to control the home environment through voice commands or smartphone apps.
Weintek manufactures human-machine interfaces HMIswhich allow people to control machine processes and monitor performance through a touchscreen interface. With AWS, AbiBird delivers a global solution that will help elderly people live independently at home. AbiBird provides a service that uses smart sensors that monitor lifestyle routines of elderly people at home without intruding on their personal privacy. LG Electronics LG launched LG ThinQ—a global brand of home appliances, consumer electronics, and services featuring artificial intelligence technology.
LG manufactures home appliances such as TVs, refrigerators, and cooling fans. The company offers products and services to enable sustainable agriculture for farmers worldwide. Using artificial intelligence, machine learning, and big data on AWS, Bigfinite helps pharmaceutical makers increase the accuracy and efficiency of their manufacturing processes while maintaining regulatory compliance.
Bigfinite develops software-as-a-service SaaS applications for industrial processes in biotech and pharma and other regulated industries. CAF Rail Services help customers reduce the life-cycle costs of its trains. Since moving to the AWS IoT environment, Centratech Systems has reduced training time for new customers from six hours to one, and has vastly expanded its customer base through an estimated 66 percent reduction in device costs.
Centratech Systems is an Australian provider of wireless monitoring and control systems used by local governments to manage water, pumping, and electricity applications. The company relies on Amazon EC2 instances to host its software used with legacy hardware, and it has recently shifted to the AWS IoT platform to manage newer, lighter, and less costly smart devices in the field.
By using AWS, Enel is saving 21 percent on compute costs and 60 percent on storage costs, has reduced provisioning time from four weeks to two days, and has transformed its business.
Enel is an Italian multinational manufacturer and distributor of electricity and gas that serves 61 million customers. EMS specializes in solutions that provide petrol retailers with performance data gathered from sensors located around petrol stations. HelloFresh, the leading company worldwide in meal-kit delivery, sends boxes loaded with delicious ingredients and matching recipes straight to your doorstep.
Founded in Berlin, the company quickly expanded internationally and currently serves over one million households in eight countries. For further developing it's IT sector, HelloFresh was looking for a best in class partner that is similarly flexible and efficient.
These services improve operational efficiency, facilitate global collaboration among employees and allow HelloFresh to better serve its customers worldwide by reducing package delivery times. IDEXX is a global leader in veterinary diagnostic equipment and reference laboratory services. Kemppi used AWS to bring its IoT solution for its flagship welding machine to market and cut the cost of software development by approximately 50 percent.
The Finnish company, with a history of innovation, designs and manufactures welding equipment and application software. Pentair increases beer filtration system performance by 10 percent, reduces costs, and increases predictability for filtration processes. The company is a global provider of water filtration systems equipped with sensors to breweries, fish farms, and other industrial and commercial customers.
People in Need is using AWS to scale an early warning system in Cambodia that alerts aboutsubscribers when floods threaten. Based in the Czech Republic, People in Need is a nongovernmental, nonprofit organization engaged in humanitarian and development work in more than 20 countries. Siemens has deployed an IoT environment at a plant in Mexico in less than eight weeks, which will help it double efficiency.
Syskron X, subsidiary of Krones AG, develops digital end-to-end solutions for production lines and uses AWS services to help customers in the food and beverage industry to optimize their supply chain.Predictive maintenance analytics captures the condition of industrial equipment so you can identify potential breakdowns before they impact production.
Businesses are constantly seeking faster ways to take advantage of the value of sensor-based information and transform it into predictive maintenance insights that people can act on quickly. Predictive maintenance insights provide valuable services, such as predicting equipment failure, real-time anomaly detection, predicting pressure spikes, and asset health monitoring.
Data scientists often develop models in a sandbox environment, then work with IT to deploy each model. Zementis leverages open industry standards, allowing users to bring both existing and new models into their AWS environments without custom coding. This framework can be used across a variety of use cases in IoT applications. AWS Marketplace is a digital catalog with thousands of software listings from independent software vendors that make it easy to find, test, buy, and deploy software that runs on AWS.
More info. Accelerate and optimize the development, deployment, and operation of smart home and smart city IoT solutions. Track data relating to human movement, location, and environment, so you can prioritize areas of focus for safety and productivity. We're here to help you get started with AWS Marketplace. Ask for or give advice on the AWS Marketplace discussion forum. Solutions in AWS Marketplace. IoT Predictive maintenance. Predictive maintenance Predictive maintenance analytics captures the condition of industrial equipment so you can identify potential breakdowns before they impact production.
Get started a Zementis from Software AG free trial. Certain messages invoke an AWS Lambda function and analytics to determine if there is an anomaly in the data that would trigger further analysis on a particular machine in a factory.
In case further analysis is required, an equipment health monitoring model or models are executed by the Zementis Server on AWS to determine what the next action should be. The results of these analytics are stored in a DynamoDB table. The equipment health monitoring IoT model deployed by Zementis invokes an AWS Lambda function that processes equipment health data during a specified time period and stores it in a DynamoDB table.
The equipment health score AWS IoT rule detects the end of a time period and invokes a Zementis deployed model that processes aggregate equipment health data to generate an equipment health score, trigger an Amazon SNS notification to the designated user sand add the score to the historic equipment health table.
AWS customers who are easily and securely connecting devices to the cloud. See more IoT solutions. Related use cases Smart home and city: device operation management. Locate, monitor, and manage connected devices at scale. Smart home and city: monitoring and response automation. Auto-create work orders and dispatch crews with smart monitoring of connected city assets. Smart home and city: accelerated IoT development.
Smart home and city: data visualization. Visualize data from your transportation fleet so you can act to maintain performance. Connected healthcare: monitoring of remote patient health applications. Track and monitor your remote patient health applications. Industrial IoT: worker safety and productivity.Prevention is better than cure.
This goes far beyond a mere Italian advertising headline from the 80s. Enel Green Power is seriously determined to prove how hydro - the oldest energy source - can stay in step with the times or even go one step ahead in terms of safety. This project showcases innovative and cutting-edge technological solutions with Enel Green Power at its forefront. EGP is the first company of Enel Group and in the world, to spearhead such a large predictive maintenance program through a public tender for manufacturers.
The program will focus on a whopping hydro power plants, the largest and most powerful in its portfolio, totalling up to 18 GW of installed capacity. PresAGHO will set forth a game-changing review of maintenance strategies, fostering a gradual switch from a preventive to a predictive type of approach.
While preventive maintenance is about constantly monitoring the status of a power plant and the proper operation of its components - and this requires adequate time and resources to perform - predictive maintenance represents instead a significant change of pace thanks to its data-based management strategy. This choice will allow Enel Green Power to further advance on the overall quality of monitoring and controls.
As a result, EGP will benefit from an optimization of performances and the bolstering of its approach on risk management, focusing on the electromechanical parts of its power plants. Such a painstaking work is leading up toduring which the ordinary preventive maintenance strategy will follow a transition roadmap towards the setting up of a predictive strategy, based on an innovative approach to failure mode and a new optimization of maintenance plans.
Maintenance strategies are to be diversified based on the requirements and sizes of each individual hydro plant. First of all, the largest types of equipment will benefit from the upgrade of existing sensors with new technologies, while the smaller types of equipment - currently lacking the most advanced sensors - will see the setup of low-cost sensors engineered and manufactured by Enel Green Power.
The strategy put in place for civil engineering works is to improve monitoring through new GIS Geographic Information System technologies and advanced algorithms, compatible with drone inspection campaigns. These new technologies, coupled with the proven experience our colleagues on managing power plants, will allow the start of the optimization process, thus shifting the selected hydro plants towards a more advanced data-driven management model.
This strategy will feature a two-fold approach, based on reference clusters. Even in the context of an energy market that requires increased levels of flexibility, this is an innovation that allows to boost performances while safeguarding sustainability. Here they are:. A fundamental step in defining the new predictive strategy will be a detailed realignment in the classification of potential faults or failure modein order to improve our data-driven approach and to measure its performance.
Moreover, Enel Green Power will partner with other utility companies to pitch for a common ground where feedback on critical situations can be sharedin order to bolster awareness on the safety of hydraulic works for electromechanical activities as well. PresAGHO will create the basis for the elaboration of predictive models addressing faults in hydroelectric power plants, while optimizing costs and standardizing maintenance procedures in various countries.
PresAGHO is a good omen for a greener and safer future. This site uses both first and third party analytics and profiling cookies to send you advertisements tailored to your personal preferences.
Quick links Where we are Join us About us Media. Enel Group Websites. Other Websites. Explore enelgreenpower. From preventive to predictive maintenance PresAGHO will set forth a game-changing review of maintenance strategies, fostering a gradual switch from a preventive to a predictive type of approach. Here they are: KDI : key diagnostic indicatorthe health status of a component KTI : key trend indicatorthe remaining time until a critical threshold is reached FPP : fault presence probabilitythe probability for the occurrence of a fault FSI : fault severity indexthe severity of the possible fault Innovation and safety A fundamental step in defining the new predictive strategy will be a detailed realignment in the classification of potential faults or failure modein order to improve our data-driven approach and to measure its performance.