Topics

Good morning, I’m Andrew Jeffery and I’m pleased to be giving you an introduction to the Maximo Application Suite.

When I’ve been asked to explain what Maximo is today and what is the Maximo Application Suite, I’ve found that going straight-in and explaining the applications in the suite is a tough starting point, as the part which most audiences will understand, Enterprise Asset Management, is only one of the suite applications. I’ve adopted an approach where we start with a little history of Maximo and explain how it has evolved to be much more than an Enterprise Asset Management system. In fact, industry analysts like IDC now talk about Asset Life-Cycle Management Applications and places IBM as having the largest market share. Starting with the history of Maximo also helps to explain the technology shift which will be needed for those clients upgrading from earlier versions of Maximo.

We’ll follow this with a slide on the Asset Management Maturity Model which helps to position some of the applications in the suite.

Most of this presentation will introduce the suite and provide a high-level overview of each of its applications. An overview of the Enterprise Asset Management application which is what traditionally people have called Maximo is a different presentation which will be found on Maximo Secrets.

Finally, we’ll look at what the IBM team are doing to ensure that Maximo continues to be recognized as a leader.

History

Did you know that Maximo is nearly 40 years old? 

It can be difficult to imagine something as large as Maximo without giving a quick preview of its heritage. In May 1985 Maximo started out as a turnkey file-based maintenance system from PSDI where the software came with an IBM PC, buying a PC was difficult then. A decade later networks, databases and Microsoft Windows were available and in 1996 PSDI launched its first client-server based Computerised Maintenance Management System (CMMS). The Maximo CMMS had assets, planning, work, inventory and procurement applications underpinned with financial transactions for interfacing to a finance system, and a rich-set of configuration capabilities.

The functional footprint of the core Maximo applications and configuration capability continued to grow and over the next decade there was an expansion of add-ons (red triangles) and industry solutions (blue triangles). PSDI had been early into Java architected solutions and by 2005 its products were entirely web architected, and browser based with a rich integration layer which was ahead of its time. The company had renamed itself MRO Software and starting with version 5 the database supported enterprise-wide asset management data structures, the organizations, sites and sets structure which some of you may be familiar with.

By 2005 you could call Maximo an Enterprise Asset Management (EAM) system, this was nearly 20 years ago. It supported a wide range of industries and asset types, facilities, manufacturing, transportation and IT including software. There was an underlying Service Management capability based on some of the IT Information Library (ITIL) v3 processes. This led MRO Software being acquired by IBM in 2006.

The diagram may look as if nothing much happened in the 5 years between 2010 and 2015, but this is untrue. The core product footprint was continuing to expand as were the add-ons and Industry Solutions. Mobile has evolved multiple times over 30 years, similarly Scheduler and Spatial have changed with the architecture and capabilities of web browsers and integration. In fact, the acquisition by IBM accelerated product development, and development is arguably at a greater level of pace today than it has ever been.

In 2015 the Maximo product was moved into IBM’s Internet of Things (IoT) division, and we started to see a set of new solutions appearing (magenta triangles) which today have evolved and are called Monitor, Health, Visual Inspection (VI), Assist and Predict. Some of these together provide capabilities for Asset Performance Management (APM). These applications use AI as a core-part of their functionalities. In the last decade two new Industry Solutions have been added for Aviation and Civil Infrastructure.

In May 2020 we started to hear about the Maximo Application Suite (MAS) and the Journey to Predict, this is the path clients need to follow to take them from routine maintenance through to predictive maintenance. MAS has a container-based architecture which will allow AI component services to be more easily consumed by the members of the suite, it is making sure that Maximo has the architecture to support itself into the future and take advantage of the wide-range of AI capabilities that IBM has developed. As you can see in the blue bar at the bottom, a software product must evolve with new technology from time to time to remain successful and relevant, MAS is one of those evolutions. It is the underlying architecture to container-based where the big difference occurs and where new technical skills will be needed for those upgrading from Maximo 7.6, the user interface and underlying database tables have hardly changed. Maximo was moved into the Maximo Application Suite in May 2021, and you will learn that it is called Maximo Manage. There have been seven incremental releases since, and the current version is MAS 8.11, it was released in September 2023. During this release, an add-on called Reliability Strategies was provided, the start of adding Reliability Centered Maintenance capabilities to Maximo.

Asset Management Maturity Model

We now know that Maximo is an Enterprise Asset Management system, an Asset Performance Management system, and it has applications supported by artificial intelligence. How are these brought together to support customers as they drive their asset maintenance strategies forward?

An Asset Management Strategy and Maturity model is used to describe where a company exists in relation to the top performing enterprises. The asset management maturity model has evolved over the years as technology has advanced, there are no longer boundaries between Enterprise Asset Management and Asset Performance Management, and we are now looking to prolong asset life and considering strategies based on Asset Lifecycle Management and how all of this can be achieved in a sustainable way.

Not all assets are the same, for some it will be quite acceptable to replace them when they have broken. Assets are becoming more instrumented and interconnected through IoT based platforms, providing near real-time asset monitoring which can be utilized to perform maintenance based on condition. But this produces large data volumes that must be turned into actionable intelligence through analytics, optimization and AI. For each asset we need to find an appropriate asset management strategy depending on its criticality to the business, that strategy may change over time as the asset approaches its end of life.

Most Computerized Maintenance Management systems will support Reactive, Run-to-Failure, Calendar and Usage based maintenance. In Maximo this uses the Preventive Maintenance application with time and/or meter-based frequencies to generate work orders and inspections, or the creation of service requests and work orders to register and fix asset defects when they occur. However, preventive maintenance is mostly ineffective at eliminating failures because failures generally occur randomly, and intrusive maintenance can also be a cause of failure. If you were performing condition-based maintenance, then preventive maintenance might only be applied to satisfy statutory frequencies or prolong manufacturer warranties.

Condition based maintenance requires inspections that record both measurements and observations, Maximo supports this. However, real-time asset condition is more effective requiring measurements to be recorded through sensors or instruments connected to SCADA, Plant/Process Information Historians, Building Management Systems (BMS) or Manufacturing Execution Systems (MES). The solution for this in the Maximo Application Suite is called Monitor. But it is impractical to monitor all assets in this way.

Risk-based maintenance is a process for selecting the target set of assets to be maintained. Assets are identified, assessed for criticality, failure modes and risk of failure and from this a maintenance plan can be formed that reduces the risk of failure. The risk assessments help you to decide which assets to monitor and what conditions to monitor. Maximo Health allows you to determine the health of your assets considering the asset’s effective age and the risk of failure in the assets operating context. The scoring methods you deploy for this can be achieved for all critical assets across the enterprise allowing you to focus effort on those assets showing poor asset health.

Predictive Maintenance requires historical failure data to train models to predict future failures. If the panacea is to perform predictive maintenance, then a starting point is to understand which assets fail and where it is worth considering predictive maintenance in the future, and on these assets start to collect good quality failure data, analysingroot cause, and the remedies taken. Start to build the failure data today that you will need for predictive maintenance tomorrow. When you have assets that fail on a regular basis then Maximo Predict can be used to predict future failure.

Financially Optimized maintenance occurs when you know when to continue to maintain an asset, when to refurbish the asset, and when to replace the asset. When you look at financially optimized maintenance holistically you are balancing risk and budget which you may do at different levels in your organization and on different sets of assets. Maximo Health has tools for both these aspects.The part of the Maximo Application Suite which most of our clients use today will only get you so far in this Asset Management Strategy and Maturity Model, to condition-based, but not real-time condition-based maintenance. The other applications in the Maximo Application Suite will need to be used to reach higher levels in this maturity model. The aspect which many forget is that these other levels require good quality data on which to make meaningful decisions, that data can be collected today, which will place you in a better position for the future.

Maximo Application Suite

Maximo Application Suite (MAS) is a set of applications providing Enterprise Asset Management (EAM), Asset Performance Management (APM) and some additional applications which are based on AI. MAS is recognized as a leading Asset Life-Cycle Management (ALM) Application.

• Maximo Manage is our EAM solution and is what most people will recognize as Maximo.
• Maximo Monitor is our Internet of Things platform providing near real-time condition monitoring.
• Maximo Health gives you insight into asset health and allows you to perform risk-based and financially optimized maintenance.
• Maximo Predict has AI-based predictive models that extend Maximo Health to predict the expected next failure date and other factors. It is the third APM application along with Monitor and Health.
• Maximo Visual Inspection uses AI to identify defects visually, through photo or video. It is used with drones as the image suggests but it could be a fixed camera on a production line.
• Maximo Assist uses AI to provide assistance to a technician, think of a search engine dedicated to your assets. It can also be used to remotely connect a technician with a human-expert, where the technician’s smart phone camera becomes the eyes for the expert where the expert can annotate the image to explain the steps to perform.

Together these applications provide the means for clients to move from EAM to Asset Lifecycle Management (ALM) which is all about extending the lifespan of assets, increasing their efficiency by reducing downtime and doing this in a sustainable way. The strategic asset lifecycle considers the asset from initial planning, through its acquisition and all the way through to its disposal and will consider not just maintenance costs but major repairs or replacement decisions and will involve risk management and investment planning. Both APM and ALM is being enabled through technology innovations, and AI.

Maximo Application Suite already uses various aspects of AI provided by tools like IBM Cloud Pak for Data, Watson Studio, Watson Machine Learning, Watson Discovery, and Watson Assistant. App Connect is used to integrate MAS with other software solutions from IBM, for example TRIRIGA and Envizi ESG Suite but also 3^rd party software solutions like Workday. There is also entitlement in Maximo Application Suite to use Cognos Analytics against the Maximo database.

Maximo Application Suite uses a container-based architecture running on IBM Red Hat OpenShift and can be deployed on premise or in a public cloud, IBM Cloud, Amazon Web Services (AWS) or Microsoft Azure. Edge devices can be used with both Maximo Monitor and Maximo Visual Inspection.

Why Maximo Application Suite (MAS)? Maximo is generally considered easy to use, and work on role-based applications is making this easier only providing the fields necessary to perform a role-based process. Enterprise Asset Management requires a core set of capabilities which are common to all industries where assets are maintained. However, there as many capabilities which are important in some industries and not others, for example management of linear assets, instrument calibration or working in a spatial context, and other capabilities which might not be considered core but are being adopted by many clients, for example, scheduling, the use of mobile devices, or adopting various ISO 55000 processes. Maximo supports all of this and extends these capabilities for certain industries.

Maximo has had the ability to configure built into the product from its very first release, it is why it was adopted by so many industries. Throughout its near 40-year life, the ability to configure has expanded with each release, this continues today with Automation Scripts and the Maximo Application Framework used to modify role-based and mobile applications. The ability to configure and customize for clients with complex assets is important, because industries are different, and processes cannot always be normalised.

Similarly, Maximo’s integration capabilities have over the years led the way, initially through necessity, Maximo never joined the Enterprise Resource Planning (ERP) world choosing to be best-of-breed and therefore needing to integrate to financial and procurement systems, when purchasing is not performed in Maximo. Since the acquisition by IBM in 2006 the integration capabilities have been partly driven by other IBM products which share the same base services as Maximo, the integration to other IBM products or its own products, like mobile and role-based applications, or the integration needed to support AI, optimization or analytic capabilities.

The Maximo Application Suite has a large footprint of capabilities all of which fall under a single license giving you entitlement to everything in the suite. There is one exception which is Maximo IT, a purchased add-on added to the suite in MAS 8.11. Licensing is a much simpler approach than we had previously where every add-on and industry solution was separately licensed, now there is a single part number, and licensing is no longer a barrier for a proof-of-concept trial.

Maximo Application Suite is based on container architecture. In comparison to virtual machines containers are lightweight and portable, they share the operating system with other containers and the image size is therefore a lot smaller than a virtual machine which contains the operating system. They are quicker to spin-up/down and use less hardware resources than the equivalent virtual machines. Containers are also easier to scale, and they provide flexibility to scale those parts which need scaling as opposed to everything that is in the virtual machine. Maximo Application Suite does require more resources than Maximo 7.6, but then the two are not-comparable, and the AI elements also require resources to run. The move to containerization is an example of the architecture change which software products need to go through from time to time as technology changes. Maximo is now set-up to take advantage of a world which will be dominated by an ever-increasing use of AI, you might say it is future-ready.For decades Maximo has been considered #1 by the industry analysts like Gartner, IDC and Arc Advisory Group, appearing in the top quadrant or arc. You’ll see later that it has the largest market share of Asset Life-Cycle Management applications according to IDC and in February 2024 it was in the Top 100 Best Software Products according to G2 who received survey responses from over 90 million people in 2023.

Maximo Manage

The first part of the Maximo Application Suite is called Maximo Manage; it is what everyone understands as Maximo. It meets the needs of Enterprise Asset Management clients, across multiple markets and asset intensive industries.

Maximo is available both on-premise or cloud based with availability on Microsoft Azure, Amazon Web Services or IBM Cloud. There are three SaaS offerings, Essentials for small businesses, Standard and Premium. What allows it to work across the enterprise is the multiple organization and multiple site data structure.

Core Maximo is a set of tightly coupled applications for managing assets, inventory, contracts, procurement, service, planning and work. There are applications to support financial integration, analytics, and wide-ranging configuration capabilities including workflow. Maximo is often customized and there are tools available to support both live changes or those requiring a maintenance window, these include migrating the changes between environments. 

With Maximo Application Suite there were changes to the software license based on AppPoints. The products which have a magenta colour are now included in Maximo; Linear, Scheduler, Calibration and a suite of mobile applications called Maximo Mobile.

Maximo can be extended with Industry Solutions for Transportation, Nuclear, Utilities, Oil and Gas, Aviation and Civil Infrastructure. These are additional sets of applications and changes to core Maximo applications to support a particular industry. The add-on products  also extend core Maximo and can be used with the Industry Solutions, but they are not specific to any one industry. 

• Health, Safety and Environment (HSE) is a large subset of the Oil and Gas Industry Solution and would be used to extend Maximo towards meeting ISO 55000 standards. 
• Service Providers is used where there are external or internal customers where the billing of work or usage/lease of assets is required. 
• Asset Configuration Manager is for the management of complex assets, particular those which may get maintained by different parties and where there is a need to understand the differences over time between as-designed and as-built, it is mainly used by aviation and train companies. 
• Spatial is the integration of ESRI’s ARCGIS in Maximo. It allows users to plan work in a spatial context. Core Maximo also has mapping capability using Google or Bing maps.
• Optimizer runs schedule optimization models developed in IBM CPLEX Studio. Models are provided for capacity planning, resource leveling, labor assignment and spatial scheduling. Optimizer is also used in other parts of the Maximo Application Suite.

Maximo IT is the most recent add-on and needs a bit of explaining. In 2006 after the IBM acquisition, a group in the IBM Tivoli division set about enhancing Maximo for IT services including those being undertaken by the Global Technical Services part of IBM on behalf of their customers. The product evolved and was until recently called IBM Control Desk (ICD). There are several Maximo clients who have both Maximo and ICD, they have always been able to coexist on the same platform and database, and they use the same configuration and integration applications. I find it easier to consider Maximo IT like an Industry Solution, a large suite of additional applications and functionality which is primarily designed for IT departments and DevOps. However, there is convergence between IT and OT (Operational Technology) the digital information being created by physical assets, and this is applicable to many industries and hence it is being considered an add-on. Bringing Maximo IT into the suite makes a lot of sense.

Maximo has a Connector for SAP including to SAP S/4HANA, and a Connector for Oracle Applications. These provide a wide range of integration points depending on the process split between Maximo and SAP/Oracle. For integration to other finance systems the Maximo Integration Framework (MIF) is used, a part of core Maximo. There is a relatively new Connector for Workday Applications, which does not yet have the same integration points as for the SAP and Oracle Connectors. This uses a product called App Connect which comes with the Maximo Application Suite. App Connect is also being used to provide both the Connector for TRIRIGA, an Integrated Workplace Management System (IWMS) used by real estate and facilities teams to manage a portfolio of building and ground assets, and the Connector for Envizi ESG Suite which pulls together Environmental, Social and Governance (ESG) data across the enterprise for analysis and reporting. Envizi supports the European Union’s CSRD rules to report sustainability disclosures. 

The Maximo Reliability Strategies add-on was also recently introduced in MAS 8.11 and is the start of Maximo’s support for Reliability Centered Maintenance (RCM). It currently consists of a single application that provides access to a large library of RCM studies across 800 asset types, and 58000 failure modes, suggesting over 5000 Preventive Maintenance activities along with the step-by-step tasks to be performed.

There are a set of Integrations to:

• Cognos Analytics 11.2.4. Maximo clients are entitled to use Cognos Analytics when used against the Maximo database, however there is a restriction to three IBM Administrative Base Users who can create/edit dashboards and reports.
• Maximo Archiving is an archiving solution for the transactional records of Maximo like work order and ticket. It uses IBM’s InfoSphere Optim Data Growth Solution which must be separately purchased.
• Environmental Intelligence Suite is a SaaS platform to monitor, predict, and respond to weather and climate impacts. It is mostly, but not exclusively used with electric utilities for outage predictions, vegetation management, renewables forecasting and alerting field technicians to weather conditions. It is also used in conjunction with Envizi and IBM Planning Analytics to provide an Enterprise Carbon Management Software solution and was recognised in the leader quadrant of Verdantix Green Quadrant: Enterprise Carbon Management Software 2023.
• MRO Inventory Optimization is a cloud platform suited to Maximo clients who wish to optimize their Maintenance Repair and Operations (MRO) inventory. It is now part of IBM Sterling Supply Chain Intelligence Suite.
• Maximo Accelerator Catalog is a way for IBM and its Business Partners to provide accelerators that extend the capability of Maximo. It was launched in MAS 8.11 with just a few products, but there is a rich ecosystem of Maximo Business Partners who have developed products over the years who may add to the catalog over time.

Maximo Monitor

Assets are becoming more complex with onboard computers and sensors, and there are real-time remote monitoring, SCADA or other types of embedded systems. Assets are generating millions of readings from thousands of data points constantly. The challenge is then how can you monitor the assets across your enterprise and identify what to focus or act on. The volume of data produced by these instruments is staggering and cannot be pulled across the network which is being used by other enterprise systems, processing must be performed close to the assets on edge devices. The analytics which filters the data and detects anomalies must also be performed at the edge with rules that forward relevant data to a central system where it can be analysed by engineers who understand the assets and can take action. When you look at the range of assets in the enterprise, they come from hundreds of manufacturers and one manufacturer may have used many different protocols over the age range of the assets that are still providing critical functions in your sites. These are some of the challenges which Maximo Monitor is addressing.

Maximo Monitor is a near real-time monitoring solution pulling data from various sources, through the IoT Tool, connections to external systems or data collectors, for example SCADA, EMS, MES or BMS, or Data Historians, from CSV files or database tables. The data can be collected in batch mode or through streaming. Calculations are performed on the data being collected and this is used in analysis to produce alerts. Analysis includes AI-based anomaly detection that detects outliers in the data, that might indicate future failure. The analysis will also identify gaps in data or flat lines, both of which may indicate a fault that needs investigating. There are more than 50 standard functions and custom functions can also be created. Custom functions on streaming data will generate an alert within a few seconds, faster than can be achieved with batch data metrics.

Maximo Monitor provides configurable dashboards, rules and alerts that can provide an enterprise-wide view of your assets. It is based on an asset hierarchy which can be synchronized with the assets in Maximo Manage through the Asset Data Dictionary (ADD) or if you are not using Maximo Manage then the asset data can be loaded. Monitoring a wide variety of assets at scale requires grouping them by device type and these can be viewed from summary dashboards. A device type is a template for a device, and may represent an asset model, whereas a device is an asset that would have a serial number. The dashboards also support the display of 3^rd party content. Alerts are configured and reported on the dashboard when a condition breaches the thresholds that you set, you can then choose to create a Service Request or Work Order in Maximo Manage.

Maximo Monitor is in part IBM’s Internet of Things platform. It uses Node-RED to connect and set-up devices, and with Node-RED you can simulate anomalies to test your alerts and dashboards. Node-Red is now open source and is widely used in the Industrial Internet of Things (IIoT) and edge computing with over 4000 connectors (source Wikipedia).

With asset data from Maximo Manage synchronized with Maximo Monitor’s device data, then metrics data from IoT devices can be used to create meter readings in Maximo Manage which can then be used by its Condition Monitoring application to automatically generate work orders when measurements exceed action limits.

To overcome the problem that a multitude of manufacturer devices with different protocols creates, the Edge Data Collector collects and unifies the data before feeding it into Maximo Monitor. A unified data solution is critical when trying to scale remote asset monitoring across the enterprise. There are a large number of pre-configured device connectors, or you can create your own. With MAS 8.11 Maximo Monitor now provides a large library of pre-configured connectors for manufacturers devices with the metrics or data points that can be tracked by those devices. This came from IBM’s acquisition of Omnio Edge. The Edge Data Collector uses a Docker Container and a gateway set up in the IoT Tool. You can also collect data from other IoT platforms like Microsoft Azure IoT.

Monitor currently only works with DB2. With streaming the analytics is performed on the data before it reaches the DB2 database.

Maximo Health

Maximo Health is the second in the Asset Performance Management (APM) applications, pulling data from Maximo Monitor and Maximo Manage but often clients use Maximo Health without implementing Maximo Monitor first and there are also examples where Maximo Health is being used standalone in advance of a clients upgrade of their Maximo 7.6 system. Maximo Health identifies the health of your assets through a scoring system that you devise. It focuses you on the right assets to maintain and when you apply this across the enterprise it reduces operational risk and reduces unnecessary preventive maintenance and its associated downtime. Maximo Health helps with capital replacement planning decisions and with MAS 8.11 there is now functionality to support Asset Investment Optimization (AIO).

Maximo Health is typically used by a reliability engineer. Most clients start using Maximo Health by defining asset health calculations for each asset class.

You start by deciding which asset and location records should be grouped together, these are called work queues. While work queues are generally aimed at identifying which asset of a type may be in poor health, they can also be used to identify which assets are missing key data that will be used in health score calculations. Custom views help to focus the work queue on the information which is relevant to it. Custom views can be personal or made public. You can launch from a record in a work queue to the asset details page where you can modify the asset or take action, create a service request, create a work order, or recalculate health scores. When creating a work order you can reference the Inspection Form that the inspector needs to perform.

The next step is to define the scoring mechanism which is normally performed by asset type and asset. This uses Maximo formulas to calculate values such as asset health, risk, criticality, end of life, and effective age. A scoring method may have multiple drivers each with multiple factors or formulas. Weighting is used to derive a percentage value and scoring ranges are used to identify, for example, good, average or poor health. Health Scoring is generally performed as a background cron task. 

Implementing Work Queues, Scoring and Dashboards is often used to replace complex spreadsheets or bespokesystems, which is why Maximo Health is implemented first using just data from Maximo Manage before adding data from Maximo Monitor later. Health scores can be improved by introducing current condition into the formulae. This can start by using inspections to collect observations and measurements, later to collect Operation Technology (OT) data derived from instruments, SCADA or other systems monitoring assets that are collected via Maximo Monitor which in turn updates Condition Monitoring measurement values which the formulas use. 

An asset’s dashboard shows a set of useful KPIs, asset information, asset scores, an asset timeline showing historical failures and work orders and future preventive maintenance activity, the operational status (current meter values), the work order maintenance history, and any replacement plans you may have created. You can create a plan to refurbish, replace or decommission an asset. 

In MAS 8.11 some of the functionality of Health and Predict Utilities was merged into Maximo Health and Maximo Predict. One aspect was colour-coded matrices which allow you to drill down to identify the assets at risk. The out of the box matrices compare criticality with one of risk, health or end of life. Custom matrices can also be created, for example to include total maintenance cost, and with Maximo Predict there is a facility for future forecasting how the matrix will look on a future date, which would allow you to take actions today.

Another aspect added to Maximo Health is the creation of investment projects. Assets belong to an asset type which in turn has a replacement plan template that is used if there is no specific replacement information for the asset. Each asset which you add to the investment plan must also contain other information like an installation date or effective age score, criticality scores and an end-of-life curve model and predicted risk curve model, these are set-up in the asset class notebook using IBM Watson Studio. A strategy is defined for the investment project, for example maintain risk, reduce risk to a certain level or stay in budget. You run the analysis using IBM Maximo Optimization and you can save the results as a scenario for side-by-side comparison. A scenario can then be submitted for approval.

The Maximo Health dashboard also now supports two new cards which were in Health and Predict Utilities:

• A Duval Triangle Card for showing the percentage of methane, ethylene and acetylene when you take Dissolved Gas Analysis samples from transformers.
• A History of Combustible Gas Analysis Card showing the history and the condition rating of each sample.

The data in Maximo Health is used by Maximo Predict. Maximo Predict is really an extension to Maximo Health and uses the same dashboards that you have already set up.

Maximo Predict

Maximo Predict extends the Maximo Health application adding Prediction Settings to the menu. There may be reliability engineers who view asset health or the results of prediction but a smaller group who run the prediction models. Maximo Predict therefore uses the same dashboards, work queues, and scoring methods.

There are five prediction models:

• Probability of Failure – uses IoT sensor and failure data to calculate a percentage risk of failure within a specified time period.
• Factors that contribute to failure – uses similar data and an analysis tree model to identify the asset attributes that are likely to cause a failure
• Predicted Failure Date – uses similar data to calculate when the next failure date will occur, allowing you to understand whether you are over or under maintaining an asset, which then provides you the evidence to adjust the asset’s maintenance schedules.
• Anomaly Detection – uses similar data to find what assets are providing measurements which seem abnormal. If you don’t have enough historic failure event data, then if an anomaly score exceeds a threshold, it will be identified.
• Asset Life Curve – uses installation date, expected decommissioning date and other asset values to estimate the probability of failure when the asset has reached a certain age.

As you can see sensor data from Maximo Monitor and previous quality failure data including failure modes from Maximo Manage are both required to progress to Maximo Predict. The predictive models need to be trained with data and this uses Watson Studio and Watson ML (Machine Learning). Historical weather data may also be needed for assets exposed to weather conditions that are likely to effect asset life.

The templates needed to help run the predictive models or modify them are provided as part of Maximo Predict, modifying the models would require data science skills. They do require a lot of data which needs to be of excellent quality otherwise the models will be skewed by the poor quality. Assets that do not fail or fail infrequently would not be good candidates for Maximo Predict unless there were many assets of the same type and model. Utility companies are good candidates for going on the journey to predictive maintenance.

Maximo Predict allows you to group assets into a predictive group where those assets can be compared from a predictive point of view. It adds a Predictions dashboard to the asset page providing the values resulting from the running of the prediction models, including trends for both failure probability and anomaly detection, and the Asset Life Curve. The asset timeline is extended to show the next predicted failure date. There are two additional work queues to show assets with a high probability of failure, and assets which are predicted to fail before the next PM.

The failure analysis tree is a diagram that builds and highlights the asset factors which are most likely to contribute to failure.

Other parts of the Health and Predict Utilities which were not merged into Maximo Health or Maximo Predict in MAS 8.11 can be downloaded from the Maximo Accelerator Catalog.

• Maximo Models for Electrical Distribution are 18 Jupyter model condition scoring templates for a variety of asset classes, transformers, circuit breakers, switchgear, and overhead/underground cables.
• Common Information Models (CIM) is a set of data used by Electrical Transmission and Distribution companies which can be loaded into Maximo

Maximo Visual Inspection

Maximo Visual Inspection (MVI) allows you to detect defects and errors visually using photo or video. It can operate from a fixed camera position, for example on a production line, or on a moving camera, for example a camera fixed to a vehicle identifying road potholes, or a drone inspecting electric transmission assets. The cameras could be those on iPhones or iPads which are running AI computer vision to detect defects or errors and acting as edge devices to reduce the data volume that might otherwise be introduced to your network, this is called Maximo Visual Inspection Mobile. If there are existing cameras, then Maximo Visual Inspection Edge, a container-based service, can be run on an edge server executing the deployed models. A fourth option, recently added, is to use Maximo Mobile Inspections to request MVI analysis and inferencing from an inspection form by uploading an image taken from the device’s camera, the augmented image will be returned from MVI.

Images are used to identify different objects; a process called labelling, and identify whether the object has a defect, or not. The AI models need to be trained with several images before they can be deployed to the edge. Maximo Visual Inspection supports high resolution images, which can identify details of an object, for example a small millimeter-sized crack on concrete bridge supports. It is used to create defects in the Maximo Civil Infrastructure industry solution, but it can also be used with Maximo Monitor to augment sensor-based anomaly detection.

The process is to label, train and then deploy the AI deep-learning models. After the model is deployed it is tested with unused images to see how it scores, a step called inferencing, but it is sometimes known as scoring. 

Models are provided for identifying the subject of an image (image classification) and items within an image (object detection), these are the only two models supported on the iOS application MVI Mobile. There are other models for image segmentation which draws a polygon around each object in the image, this can be used to measure the size or volume, for example the size of a dent on a car, or the volume of a pothole. The anomaly detection model is where training is based on many good images and a few images with a defect, this is useful when defect images are in short supply. Images can be pipelined so that inference results from one model can be used as the input for another model, a composite model. For a video, the labeling is performed on each frame.

One of the issues with image-based AI is that a large and varied data set may not be available. Maximo Visual Inspection has image augmentation capabilities which create additional images applying operations like flip, blur, sharpen and crop. This improves model accuracy faster. After the model is deployed there is an auto-label function which can be applied to new images which can then be added to the data set, and the model retrained with these additional images to improve the accuracy of the model. This auto-labelling function speeds up the deployment process, and it can be applied iteratively.

During training Maximo Visual Inspection will provide basic feedback on accuracy, but there is also advanced feedback commonly used by data scientists including a precision recall curve and confusion matrix.

When deployed to Maximo Visual Inspection Mobile there is a choice to run the models remotely or the models can be exported to allow inferencing to occur on the device while offline. MVI Mobile can send alerts to Maximo Visual Inspection based on the business rules you configure. Images captured on the iOS device can also be used for model retraining after loading them into Maximo Visual Inspection. 

Maximo Visual Inspection Edge allows multiple cameras to be attached to the same edge server, for example multiple cameras on a production line. It also allows special purpose cameras to be used, such as high-speed, thermographic or magnification cameras. A wizard is used to set-up Maximo Visual Inspection Edge on the edge server including configuring the alert messages, for example send an inspection failed alert to the quality management team if a broken part has been identified with >90% confidence. Alerting uses a MQTT broker and the Twilio messaging service, a Twilio account will be needed. MQTT messages can also be used to perform real-time actions on connected machines that also support MQTT messages. Dashboards can be used to create a summarized view of inspections passing, failing, or inconclusive, the number of alerts raised from the various camera stations being controlled by Maximo Visual Inspection Edge.

Maximo Visual Inspection is easy to use and can be setup and trained by a reliability engineer, it does not require a data scientist. The solution can scale globally, and it is designed to continuously improve model accuracy. It has the advantage of being able to work 24/7, producing consistent results. Human inspection requires breaks, and consistency deteriorates with the onset of tiredness.

Did you know that the first autonomous ship to cross the Atlantic ocean, the Mayflower, used six on-board cameras to assist with navigation and to recognize objects in the sea. This used the same visual AI technology as that being used with Maximo Visual Inspection. 

Maximo Assist

An aging workforce is leading to a loss of tribal knowledge and the volume of both structured and unstructured data is making it difficult to find the right data to solve a problem. Clients are finding that more technician time is being spent researching solutions to asset failures than the time it is taking to fix those problems. However, the good news is that there is more digital information becoming available through connected devices or digital twins and AI technology is being embedded in software as it is with Maximo Assist.

Today’s workforce is technically savvy. Pen and paper have made way to mobile applications, but there is a lot of technology out there that can still be harnessed to support today’s technicians, including AI based assistance, augmented reality, conversational user interfaces, and technology supporting safety at the workplace like wearables and beacon technology. The organizations that are taking advantage of this new technology are sharing expertise and knowledge between technicians to improve first time fix rates and reduce mean time to repair.

Maximo Assist has been integrated into Maximo Mobile and empowers your technicians in three ways as the slide shows. From left to right they are, AI guidance made with natural language queries, a diagnostic service for troubleshooting, and collaboration with an expert who may be in a remote location.

The AI guidance requires relevant data to be ingested which you do using the Maximo Assist Studio. This includes Root Cause Failure Analysis (RCFA) and Failure Modes Effects Analysis (FMEA), engineering manuals, manufacturer and owner manuals, asset data and work order history, basically anything which technicians use to solve problems on their assets is relevant. This is the AI knowledge base which Maximo Assist uses. When Maximo Assist is launched on the mobile device the technician can ask a question using natural language as they would with a search engine and Maximo Assist will provide the results with a percentage level of confidence, without, of course, the adverts or sponsored links. The relevant text is highlighted along with the title of the source document and the technician can drill into the document to see whether it provides the answer they were looking for, marking the result as useful, or not.

At the top of the first screenshot on the left are two buttons Diagnose and Collaborate. The Diagnose feature asks a set of troubleshooting questions and based on the answers given leads the technician towards a likely solution. The diagnosis model would be created by your expert, likely a reliability engineer who is using Maximo Assist Studio. There are three diagnosis model types to choose from. The rules-based model is most suited to use with existing Root Cause Failure Analysis (RCFA) and Failure Modes Effects Analysis (FMEA) documentation. A Decision Tree model can use historical and feedback data as the starting point, trained models then present this as a series of questions expecting binary responses that direct the technician to the next question to answer. The third type is a Bayesian network model which can provide multiple solutions with the associated probabilities that it is the best solution. This model is used when there are multiple symptoms and multiple possible causes of those symptoms. The reliability engineer will need to train and refine the models before they are used and will continue to refine the model over time based on real case feedback.

The Collaborate feature uses augmented-reality and connects the technician with an available expert based on their area of expertise. Once connected, the technician’s camera is used as the eyes of the expert, and the expert can apply tags and annotate what they are seeing to guide the remote technician. The session history and chat is saved on the work order for future reference and/or added to the AI knowledge base so that it can be found when a similar query is made in the future. The screenshot on the right is showing the collaboration summary at the end of a collaboration session.

Maximo Assist Studio is where you manage the documents used and where you build, train and refine the diagnosis models, this tool can be used by a Reliability Engineer, it does not need a Data Scientist. Maximo Assist is dependent on the ingested data, and you could start by collecting relevant documentation and data regarding the assets you consider highest priority for testing the benefits of Maximo Assist. You could also ask experienced technicians to record videos of maintenance activities on these assets. This is a great task for an apprentice/junior as they are out in the field getting to know the assets with their mentor. It is likely the videos will need to be edited to cut out unnecessary time, so that they are short but relevant, before they are then loaded into Maximo Assist.

Preventive maintenance routines tend to be well documented; it is what to do when failures occur which is where Maximo Assist will provide the most benefits. The Root Cause Failure Analysis and Failure Modes & Effects Analysis are key documents for Maximo Assist, as are documenting typical failure repair steps which can also be added as Job Plans.

Maximo Assist works best when the technician is online, and you must be online to use the collaboration tool. A workaround for the query and diagnose function is to download useful documents while online. If you are going to do this then make sure that the documents you load into Maximo Assist will be readable on the mobile device after they have been downloaded. With a bit of co-ordination, you can split the document loading and model training between multiple teams by having each team focus on a single asset type, and for other teams to provide feedback during real case testing. Maximo Assist is available in 22 languages, but you should be aware that not all languages are supported in the query feature. Maximo Assist uses IBM Watson Discovery 4.5.

Maximo continues to lead

Maximo continues to maintain a leadership position. For many years, decades even, industry analysts like Gartner, IDC and ARC Advisory Group made Maximo the leader in Enterprise Asset Management. Now the same analysts are looking at Asset Life-Cycle Management and in November 2023 IDC suggested that IBM Maximo Application Suite had more than 10% market share. As you can see from the pie chart there are many other systems that make up nearly two-thirds of the market share. Its competitors tend to be those where there are solutions that are part of an ERP system. Maximo is not an ERP system; it is a best of breed asset management software product which can interface with ERP or other financial systems.

In February 2024 G2 announced Maximo Application Suite in the top 100 of all software products, this was from over 90 million survey responses during 2023. Also, in February 2024 IFI CLAIMS published a study on Generative AI patents which amounts to only 22% of the total AI patents. It placed IBM a long way ahead of other companies in this space. “IBM is the elephant dancing in this AI boom, with 1,591 applications, a third more than Google.” ChatGPT which was the cause of the AI boom in social discussions isn’t in the top 25 of companies with Generative AI patents. Fortunately for us IBM has been working across AI for many years, and no doubt it has a lot of patents in other AI spaces as well, the remaining 78%.

G2 Software Award

IFI CLAIMS study into generative AI

Maximo Next Release

2024 looks to be an exciting year with the expected release of Maximo Application Suite 9.0, which is targeted for June 2024. While there seems to be a lot going on IBM has said that it is an incremental release, and certainly not a major release as a .0 implies. The reason for calling it MAS 9.0 is to align the version numbers of each of the products. We will need to wait to find out what the next release contains, things can change. I’ll update this page when it has been released.

The investment themes which seem to characterise where IBM are working, and these are my observations, are:

• Asset Life-Cycle Management, new functionality like Reliability, and Asset Investment Optimisation are likely to see continued development. There is also likely to be a Sustainability footprint being developed in Maximo to support the reporting performed in Envizi.
• Mobile and User Interface, there is some finishing off to completely replace Maximo Anywhere and the Work Centers, but there are obvious opportunities for mobile applications supporting the industry solutions and add-ons like HSE, ACM, and role-based applications and dashboards will continue to be developed. The Maximo Application Framework still has some way to go to get to the same level as Application Designer.
• Field Service Management incorporates Scheduler, (and Optimizer), Spatial and the Technician mobile application, some of the most widely used areas of Maximo that are not the core modules. Some of the Scheduler applications will need a UI overhaul and this has started with the Scheduling and Dispatching dashboards, Linear is not yet part of Mobile Technician, and a lot of existing clients would like to venture into schedule optimization, and I expect the capability in this area will be strengthened (hope this is not wishful thinking on my part).
• IT/OT Convergence. ICD has recently been brought into the suite as Maximo IT to support this initiative and the first step is it needs to be made compatible with the other Industry Solutions and add-ons, for example you can’t currently use HSE/O&G with it. I would think there would be something to tie Maximo Monitor to the Maximo IT processes, for example firmware updates, change management perhaps. Maximo Monitor is a subset of the IoT platform and perhaps there are other parts to be brought across now that the IoT platform no longer exists as its own product.
• AI is fuelling the rise in the IBM share price, and it has a wealth of capability with watsonx. I think it obvious that additional AI capabilities will be added to Maximo. The areas of Maximo Application Suite which already utilise AI will continue to be strengthened, particularly Visual Inspection, Health and Predict.

From the published roadmap for Maximo Application Suite 9.0 we can expect to see:

Maximo Manage (EAM)

• RCM – Reliability Strategies was a new application in MAS 8.11 and there is expected to be capability to create your own strategy (Builder)
• Scheduler – The Scheduling and Dispatching Dashboards are quite recent additions and continued development is expected.
• Spatial – Performance improvements and standardisation of UI between desktop and mobile
• Support for Cognos 12

Mobility and User interface (UI)

• Mobile – Issues and Transfers, continued development of the application to support transfers including staging and shipment. There has been no reference to picking which was in the Manage Inventory Work Center.
• Mobile – Calibration, it has been on previous roadmaps
• Mobile – Miscellaneous enhancements across other applications is expected
• UI – Inspection Form Builder, to replace the Work Center
• UI – Operational Dashboard, Work Orders and Work Queues are all relatively new applications and continued development is expected
• UI – Continued development of the Maximo Application Framework (MAF) 

APM, and AI

• Visual Inspection – Support for GigE Vision industrial cameras. Image blurring outside of detection areas, management of old images/videos are expected.
• Monitor – Some user interface redesign. Add TRIRIGA use cases into Monitor for CISCO occupancy sensors.
• Health – Mean Time Between Failures. Identify missing asset data. Provide OOTB scoring calculations.AI – Work order AI around failure codes and work order approval

Maximo 2024 Roadmap

Recap

I hope you enjoyed the presentation and now have an appreciation of the scope of Maximo Application Suite. 

To summarise:

• Maximo has a rich history spanning nearly 40 years
• It is more than an Enterprise Asset Management system, it has an IoT platform, Asset Performance Management and AI applications
• Maximo is recognized as a leader is Asset Lifecycle Management, and IBM is a leader in AI
• The future is looking bright

Thank you for listening