In 2021 I performed design visioning to achieve Asset Data Unification in the Maximo Application Suite and drive a common terminology, asset data dictionary, and UX experience strategy for the Maximo suite applications integration use-cases.
Objective
We were a collection of applications loosely integrated and intently focused on addressing an aspect of optimizing an asset. We are now a unified solution that optimizes an asset using the data you have and makes that available everywhere across the asset lifecycle. We are a Maximo application suite.
The objective of this information modeling aimed to identify the key data integration use cases across the applications on the Maximo site.
The outcome of the discovery was driving the information model schema for a Data Dictionary backend service, using the KITT semantic graph service. The discovery also defined concepts for unifying a common integration experience.
Maximo application suite
Maximo application suite provides intelligent asset management through a set of applications for asset monitoring, asset management, predictive maintenance, and reliability planning.
The suite is a single, integrated cloud-based platform that uses AI, IoT, and analytics to optimize performance, extend asset lifecycles, and reduce operational downtime and costs.
The suite includes Manage, Monitor, Health, Predict, and Mobile applications.
Learn more about the IBM Maximo Application Suite on IBM.com.
Maximo application information models
The application information model, defining the foundational objects and their relations was identified in discovery sessions with each Maximo suite application team. Key objects, their workflows, and use cases were discussed to define common suite-level object types, dependencies, and integration requirements.
Maximo Manage
Maximo Manage, also known as Maximo EAM, has a rich information model. The core model defined for a data dictionary includes sites, systems, locations, assets, meters, work orders, and failure code objects.
The Maximo Manage information model.
- A production Site has operational Systems.
- A System has a hierarchy of functional Locations.
- An Asset, and its Asset and Item parts, are installed at a functional Location.
- Assets and Locations have Meters.
- Meter readings are provided as Data from Devices.
- A Work order is related to an Asset and may be classified by a Failure code.
The Maximo Manage object relationship model.
Maximo Monitor
Maximo Monitor, also including the IoT platform, has a rich device, data, function, and alert model. The Monitor model also associates devices with the assets as data sources. In Maximo Manage, the IoT Platform registers and connects devices, and ingests time-series device data. Maximo Monitor provides a streaming architecture that processes ingested time-series data using statistical functions and AI models.
The Maximo Monitor information model.
- An Asset is of an Asset Type.
- An Asset is installed at a Location.
- An Asset has Data metrics that may be Raw, Computed, or Alerts.
- Raw metrics are ingested from physically connected devices.
- Computed metrics use KPI Functions to Transform, Aggregate, or detect Alerts, using raw metrics data as parameters.
- Asset data is monitored on Dashboards.
- Assets of a common asset type may share a common Dashboard instance layout.
IoT Platform information model.
- A Device is of a Device Type
- Data from a connected device is received as Events of an Event Type.
- A Device has private data Properties in the Physical Interface that mapped to Events
- A Device as public data Properties in the Logical Interface that is mapped to the Physical Interface Properties.
- A Raw Metric in an Asset receives data from a Device Property.
Integration points
- Assets are logically equivalent to Assets in Manage.
- Locations are logically equivalent to Locations in Manage.
- Alerts are logically related to Service Requests and Work orders in Manage.
The Maximo Monitor object relationship model.
Maximo Health and Predict
Maximo Health and Predict is providing conditional, preventive, and predictive maintenance data for Assets. The current condition of an asset is defined by the asset’s historical data, computed as a health score. A future asset condition is defined by failure predictions and recommended maintenance actions.
Maximo Health and Predict information model.
- An Asset may be part of a Health Scoring Group, and a Predict Group.
- A Health Score is computed by a Health Scoring Function that uses Asset Mater Data and Asset Metadata.
- A prediction is computed by a Prediction Model, that provides predicted Asset Details and recommended Actions, such as creating a Work order, a Service request, or a Replacement plan.
Integration points
- Recommended actions are Manage work order actions.
- Prediction models use Asset Meter data in Manage, and Asset Device data in Monitor.
The Maximo Health and Predict object relationship model.
Maximo suite model
A common Maximo suite model was derived from the abstract Maximo suite terminology and common object types from the Maximo suite application information models.
- System
- Location
- Asset | Asset Group
- Data
- Function
- Alert
- Work order
In the diagram below, the abstract object types, shown in grey, and the relationships to concrete application object types.
- Assets are defined as a parts hierarchy.
- Assets are related to the system location hierarchy.
- Data is unifying multiple types and sources.
- Functions unify computational formulas and AI models.
- Assets are grouped for comparative health scores and predictions.
- Alerts on abnormal asset Data generate maintenance Work orders.
The Maximo Application Suite object relationship model.
Maximo Data Dictionary
Digital Twin Semantic Knowledge Graph
The Digital Twin Semantic Knowledge Graph technologies, developed for Industry 4.0 by IBM Research, is implementing a Maximo Data Dictionary service in the Maximo suite.
The Semantic Knowledge Graph service is designed to address many problems in data integration and management.
- Providing a semantic model that simplifies data modeling by implementing a semantic representation, easily extended and refined to specific industry use case requirements.
- Allows users and developers to view and filter data to their needs.
- Enables AI and machine learning at scale by providing machine learning algorithms good understanding of the context of the analytic subject.
- Join data from different services, multi-cloud, and multi-tenant sources.
- Optimized for performance around solving complex relational queries on data to enable powerful queries in real time.
Semantic graph models describing industry data are more expressive and flexible than traditional data models. The Semantic Knowledge Graph service is a cloud solution that combines an extensible semantic model, with a data platform to enable easy access to data and to drive analytics at scale. It’s neither a graph database nor a relational database. It uses a graph to semantically describe and connect data that may be stored in relational databases, NoSQL databases, object stores, or data lakes. The Maximo suite creates a unified suite access layer that provides search and queries to common data and links to each application data store for object data details.
The Maximo Data Dictionary consists of a semantic core model. The model is kept simple and abstract to model the core of data that should be exchanged by the applications. Applications can specify extensions to the model to map and extend this core model to their application data models.
Learn more about IBM Digital Twin Semantic Knowledge Graph technologies and the Maximo Data Dictionary.
Scaling Knowledge Graphs for Automating AI of Digital Twins.
Maximo Data Dictionary schema
The Data Dictionary schema.
- Node elements. The asset and location hierarchy is designed using a semantic graph and is defined using a generic super-type Node representing Location and Asset. A node can contain other nodes using the hasPart relationship and relate via relates to other nodes.
- Data elements. Data represents time-series data associated with assets or locations. Data may have relations to other data sets such as cleaned data may refer to raw data.
- Task elements. A task represents work such as a work order, ticket, or service request.
- Organization elements. An organization is a record that identifies a unique legal entity. The data set for an organization includes information that companies or other distinct legal entities might share, such as calendars, vendors, and financial information. An organization can contain one or more sites.
- UI elements represent Dashboards and Cards on dashboards.
The data dictionary core schema is extended and mapped to specific application object types.
Data Dictionary Asset mappings.
Data Dictionary Location mappings.
Data Dictionary Data mappings.
Data Dictionary Work Order mappings.
Data Dictionary Function mappings.
Data Dictionary Organization mappings.
Data Dictionary Dashboard mappings.
Maximo Data Dictionary integrations
The Asset Data Dictionary is installed with the Maximo suite and manages a repository of data and metadata that facilitates data sharing and integration between Maximo Manage and other applications. Using pre-defined artifacts in the Maximo Manage integration framework, data in Maximo Manage can be synchronized with the Asset Data Dictionary, which in turn integrates the data into other applications, such as Maximo Monitor.
Pre-defined artifacts are available for integrating data about organizations, sites, assets, asset hierarchies, operating locations, location systems, and location hierarchies.
The Asset Data Dictionary provides seamless adoption across Maximo Manage and Maximo Monitor and allows organizations to adopt paths.
Manage first. Many customers are using Maximo EAM in operations and have established use of location and assets. When adopting Maximo Monitor, the data dictionary synchronizes the system hierarchy and provides assets and meters as integration points for instrumented devices and real-time sensor data.
Monitor first. Customers starting to use Monitor can create a system hierarchy with devices, assets, locations, and dashboards. When adopting formal enterprise asset management, the location hierarchy and assets and created in Maximo Manager as managed objects.
The Maximo Monitor to Maximo Manage integration is a one-time process. The data synchronization can be performed manually when needed, at regularly scheduled intervals, or as event-based updates when data changes.
Learn more about Integrating with Asset Data Dictionary.