- SAP Managed Tags:
At SAP we aim to continuously improve and evolve the data model to better suit our industry and customer requirements. As such we have worked with many customers and industry teams to define and develop the Indicator concept. Design started in 2014 and development of the indicator concept began in 2015 as part of Asset Central Foundation, Asset Intelligence Network, Predictive Maintenance, Asset Strategy & Performance Management (ASPM) and various other partner and SAP solutions.
Indicators indicate the past, current and future state or level of something. An indicator can be either a measured or calculated or assessed value. The indicator value can come from a live sensor data stream, be manually input as part of an inspection, be calculated as part of an assessment or received from a manufacturer as part of the handover of asset information. So:
hence we need Indicators!
Indicators are created in the "template" application and are grouped into Indicator Groups. Those indicator groups can then be used as part of one or more Model or Equipment templates. Importantly the indicators are inherited from the model to the equipment, so that you can define your indicators once at model (i.e. product) level and have them appear on all equipment automatically. You can then fine-tune them at equipment level.
The data model for an individual indicator is as follows:
The rules engine consume indicator data from the time-series service to evaluate the condition of an asset. Multiple indicator values can be combined with thresholds as well as logic operators for master data values. The output of the rule evaluation is an automatically created alerts, notification and / or emails.
Equipment Indicator Charts consume indicator data from the time-series service and can be visualized at various timescales. They are visible on the equipment page and also on the work order and notification page. The following chart shows 4 indicators for a single equipment with an overlay of alerts / alarms.
In addition to alerts the charts can also be overlaid with Notification and Work Order event information, as well as thresholds (for condition monitoring) and can be setup to refresh on a regular time interval (i.e. every minute) for display on a large monitor in a control room environment. The settings for the chart are controlled via the following UI:
Indicator values for multiple equipment can also be visualized in a single chart - the below shows indicators for 4 different equipment and the current thresholds of one of those indicators which is then used in condition monitoring:
This analysis tool visualizes the Indicator values for one or more equipment in a tabular format along with attributes and other equipment header values. The following shows a mix of pump attributes (max pressure), pump current indicators values (bearing temperature and health score) and pump header attributes (location and model) all in one configurable table:
Historical data of the Indicator is used to train the machine anomaly and failure predictive learning models. This results in an update of Probability of Failure (PoF) and / or Health Score global indicators for the equipment
Forecasting of the possible Indicator values based on the existing data set to determine when an indicator will in the future exceed a threshold. This drives advance maintenance notification creation
With machine learning the system proactively identifies which indicators and indicator values best correlate with historical failures. Recommendations for explicit rules for condition monitoring are then proposed which can be passed into the rules engine. This helps the end user identify patterns to automatically setup failure mode driven condition monitoring.
Currently shows all the Indicators with their values for one individual equipment or model. From here the user can configure thresholds and also configure IoT Synchronization
Upon completion of the Risk assessment process the global indicators for Risk, Normalized Risk and Criticality are updated. When completing an FMEA process the Risk Priority Number (RPN) is also updated
A key part of managing the performance of an asset is to perform regular inspections or "checklists". This can be performed initially by the OEM as a Factory Acceptance Test (shared with the operator via the network) and then at a certain interval be performed by the operator. The checklist process captures input from the user to update the indicators for one or more equipment. This history of who did the check, their role, additional documentation forms a complete record. This can then subsequently trigger the condition monitoring (based on thresholds) or machine learning functions
Business parties can share the Indicator values on the Network with their connected parties. For instance a manufacturer can share the expected mean time between failure to an operator. Or an operator can share time-series 2D charts with their invited business partners. Sharing is always optional and is by default not enabled unless the an authorized user first connects to the business partner and then chooses to share the values.
Multiple partner application already consume and update the indicators. A good example is the AsInt solution for Risk Based Inspections which directly uses and updates the indicators from complex RBI calculations on the SAP Cloud Platform. Developers can view the following documentation in the APIHub:
One special area to mention is that SAP has standardized on a certain set of indicators that will be used across all assets that are managed in our public cloud. This drives standardization within and across industries and is particularly important for network use-cases where information needs to be shared between multiple parties.
Examples of global indicators include Remaining Useful Life (RUL), Operating Hours or Mean Time Between Failure (MTBF) and the list can be found in the template application using the filter:
Global Indicators are published by SAP and available in all accounts. This set of indicators are recommended for use by all customers. Please note the metadata cannot be edited and the PdMS and ASPM applications expect these indicators to be present as part of their normal processing.
Indicators can be viewed offline and online in the native Android and Apple mobile Asset Manager application. Users in the field can view the recommended threshold from the manufacturer or the latest predicted Remaining useful life of an asset when servicing the equipment. This extra insight into documentation or health and life insights often will result in direct time and work accuracy savings for the technician.
Indicators represent a significant evolution for Asset Intensive customers. They are network enabled for collaboration and IoT enabled for connection with time-series services. Indicators allow you to perform enterprise wide condition monitoring and evolve into more advanced predictive algorithms over time. This is one of the key foundations for predictive maintenance and asset strategy topics within the SAP Asset Management suite.
Why do we choose the term Indicator?
Indicators indicate the past, current and future state or level of something. An indicator can be either a measured or calculated or assessed value. The indicator value can come from a live sensor data stream, be manually input as part of an inspection, be calculated as part of an assessment or received from a manufacturer as part of the handover of asset information. So:
- We need more than just a point where a measurement occurs
- But it is not always a Key or Performance related indicator (i.e. it is often less than a KPI)
hence we need Indicators!
What are some of the use case for indicators?
- As a Maintenance Technician I want to view historically time-series data about the machine to understand how the machine has been operating to help with diagnosis (i.e. has oil pressure been increasing)
- As a Reliability Engineer I want to define, model and visualize life and health indicators so that I can monitor the health (i.e. current and future condition) and life (i.e. how old, how many cycles) status of an equipment
- As a Reliability Engineer I want to define, model and visualize statistical reliability, availability and maintainability indicators so that I can consider them in defining the optimal maintenance strategy of my equipment
- As a Process Engineer I want to define, model and visualize production indicators so that I can monitor the production status of an equipment/location/system.
- As a Data Scientist or Engineer I want to define and model inputs and outputs of my Machine Learning or Simulation algorithm so that I can consume them in the Machine Learning Engine
- As a Manufacturer / OEM I want to publish and share the recommended thresholds or operating windows for various indicators (i.e. air pressure) that my customers should operate my products within
What is the data model of the indicators?
Indicators are created in the "template" application and are grouped into Indicator Groups. Those indicator groups can then be used as part of one or more Model or Equipment templates. Importantly the indicators are inherited from the model to the equipment, so that you can define your indicators once at model (i.e. product) level and have them appear on all equipment automatically. You can then fine-tune them at equipment level.
The data model for an individual indicator is as follows:
Fourteen ways that indicators are already used in Intelligent Asset Management
1) Predictive Maintenance & Service - Rules:
The rules engine consume indicator data from the time-series service to evaluate the condition of an asset. Multiple indicator values can be combined with thresholds as well as logic operators for master data values. The output of the rule evaluation is an automatically created alerts, notification and / or emails.
2) Predictive Maintenance & Service - Equipment Time-series Charting
Equipment Indicator Charts consume indicator data from the time-series service and can be visualized at various timescales. They are visible on the equipment page and also on the work order and notification page. The following chart shows 4 indicators for a single equipment with an overlay of alerts / alarms.
In addition to alerts the charts can also be overlaid with Notification and Work Order event information, as well as thresholds (for condition monitoring) and can be setup to refresh on a regular time interval (i.e. every minute) for display on a large monitor in a control room environment. The settings for the chart are controlled via the following UI:3) Predictive Maintenance & Service - Fleet or Multiple Equipment Time-series Charting
Indicator values for multiple equipment can also be visualized in a single chart - the below shows indicators for 4 different equipment and the current thresholds of one of those indicators which is then used in condition monitoring:
4) Predictive Maintenance & Service - Equipment Indicator Analysis Tool:
This analysis tool visualizes the Indicator values for one or more equipment in a tabular format along with attributes and other equipment header values. The following shows a mix of pump attributes (max pressure), pump current indicators values (bearing temperature and health score) and pump header attributes (location and model) all in one configurable table:
5) Predictive Maintenance & Service - Machine Learning
Historical data of the Indicator is used to train the machine anomaly and failure predictive learning models. This results in an update of Probability of Failure (PoF) and / or Health Score global indicators for the equipment
6) Predictive Maintenance & Service - Indicator Forecasting
Forecasting of the possible Indicator values based on the existing data set to determine when an indicator will in the future exceed a threshold. This drives advance maintenance notification creation
7) Predictive Maintenance & Service - Leading Indicators
With machine learning the system proactively identifies which indicators and indicator values best correlate with historical failures. Recommendations for explicit rules for condition monitoring are then proposed which can be passed into the rules engine. This helps the end user identify patterns to automatically setup failure mode driven condition monitoring.
😎 Asset Central Foundation - Equipment & Model Indicator List
Currently shows all the Indicators with their values for one individual equipment or model. From here the user can configure thresholds and also configure IoT Synchronization
9) Asset Strategy & Performance - Risk & Criticality & FMEA
Upon completion of the Risk assessment process the global indicators for Risk, Normalized Risk and Criticality are updated. When completing an FMEA process the Risk Priority Number (RPN) is also updated
10) Asset Strategy & Performance - Checklist and Inspection
A key part of managing the performance of an asset is to perform regular inspections or "checklists". This can be performed initially by the OEM as a Factory Acceptance Test (shared with the operator via the network) and then at a certain interval be performed by the operator. The checklist process captures input from the user to update the indicators for one or more equipment. This history of who did the check, their role, additional documentation forms a complete record. This can then subsequently trigger the condition monitoring (based on thresholds) or machine learning functions
11) Network & Collaboration
Business parties can share the Indicator values on the Network with their connected parties. For instance a manufacturer can share the expected mean time between failure to an operator. Or an operator can share time-series 2D charts with their invited business partners. Sharing is always optional and is by default not enabled unless the an authorized user first connects to the business partner and then chooses to share the values.
12) Partner and Customer Extensions and Applications
Multiple partner application already consume and update the indicators. A good example is the AsInt solution for Risk Based Inspections which directly uses and updates the indicators from complex RBI calculations on the SAP Cloud Platform. Developers can view the following documentation in the APIHub:
13) Global Indicators
One special area to mention is that SAP has standardized on a certain set of indicators that will be used across all assets that are managed in our public cloud. This drives standardization within and across industries and is particularly important for network use-cases where information needs to be shared between multiple parties.
Examples of global indicators include Remaining Useful Life (RUL), Operating Hours or Mean Time Between Failure (MTBF) and the list can be found in the template application using the filter:
Global Indicators are published by SAP and available in all accounts. This set of indicators are recommended for use by all customers. Please note the metadata cannot be edited and the PdMS and ASPM applications expect these indicators to be present as part of their normal processing.
14) Make it mobile with Asset Manager
Indicators can be viewed offline and online in the native Android and Apple mobile Asset Manager application. Users in the field can view the recommended threshold from the manufacturer or the latest predicted Remaining useful life of an asset when servicing the equipment. This extra insight into documentation or health and life insights often will result in direct time and work accuracy savings for the technician.
Summary
Indicators represent a significant evolution for Asset Intensive customers. They are network enabled for collaboration and IoT enabled for connection with time-series services. Indicators allow you to perform enterprise wide condition monitoring and evolve into more advanced predictive algorithms over time. This is one of the key foundations for predictive maintenance and asset strategy topics within the SAP Asset Management suite.
This is from an ongoing series of blogs covering top topics for asset management practitioners and experts. The other blogs in the series are:
|
| User | Count |
|---|---|
| 9 | |
| 6 | |
| 5 | |
| 4 | |
| 3 | |
| 2 | |
| 2 | |
| 2 | |
| 1 | |
| 1 |
