Predix Asset Performance Management
In 2015, GE acquired Alstom’s power and grid businesses. GE’s Grid Solutions business serves customers globally with over 16,500 employees in approximately 80 countries. In today’s age of accelerating energy transition, the reliability, security, and efficiency of the grid remain paramount. Grid Solutions helps utilities and industry to effectively manage electricity from the point of generation to the point of consumption, maximizing the reliability, security, and efficiency of the grid.
A need for an extensible software platform
While acquisitions are typically beneficial for all parties involved, they also come with a unique set of challenges. The challenge, in this case, was to recreate a successful software application on an extensible software platform where the platform and application benefitted and complimented each other.
Following the acquisition of Alstom, an initiative was started with the goal of finding a replacement for a legacy solution called e-Terraassetcare (ETAC). Grid Solutions found a suitable replacement with Predix Asset Performance Management (Predix APM). A cross-functional team was formed across the United States, Canada, and France to support this initiative. These resources followed software development best practices of SCRUM/Agile, and worked collaboratively using a common application platform—Predix APM—to build the ETAC replacement. The new solution was named Grid APM since it is specific to electrical grid assets. However, the methodology is more broadly applicable to a wide array of industrial assets.
The development of asset blueprints
In Grid APM, a series of blueprints were created to model the behavior and performance of grid assets. These blueprinrts are software representations of assets and processes that are used to understand, predict, and optimize performance in order to achieve improved business outcomes. Providing the owner/operator with this asset-specific content decreases software deployment and implementation time and cost by as much as 80%. The shift of “content creation” from the owner/operator to the OEM greatly decreases the time-to-value for the owner/operator. When the blueprint content is applied to physical assets and is fed with the proper data associated with operation and maintenance of those assets, a performance-based Asset Health Map is created. This map provides the end user with a holistic view of the health and performance of the physical assets within their fleet, enabling condition based maintenance planning, maintenance and asset replacement prioritization, cost optimization, reliability improvement, and increased safety. When blueprints have been deployed to physical assets, the data generated by those physical assets creates a “digital twin”.
Scaling the Grid APM capabilities to accommodate hundreds of thousands of assets and the data being generated by the assets during operation and maintenance in a cloud environment is the core of Predix APM. At this macro scale, the user has the ability to assess the health of systems (such as functional collection of blueprints) or groups of similar assets (such as a fleet). This enables the comparison of performance and strategy effectiveness.
The new Grid APM solution follows the well-known “Plan, Do, Check, Act” (PDCA) model, where the risk/criticality of the asset is assessed, failure modes are identified, mitigating actions are put in place, and conditional assessment data is gathered. The conditional assessment data which is gathered includes; inspection results, lab analyses, thermography, alerts generated by real-time analytics, and more. Taking PDCA “full-cycle” in a software application is a capability that is unique to Predix APM. In this case, each risk is associated with one or many actions. When the action is executed some data is generated, and the data that’s generated during action execution is the digital thread that binds the health of the asset to the risk that the action is mitigating. With the deep industrial domain expertise of Grid Solutions, the team was also able to include condition-based maintenance recommendations as part of the Grid APM solution that provide the owner/operator a list of condition-based maintenance actions to be integrated in his maintenance plan.
The team was able to use many core or foundational Predix APM elements to complete its Grid APM solution. The capabilities that are featured prominently and used over and again during the development include:
- Asset strategy manager was used to implement and manage actions which correlate to risks.
- Failure modes and effects analysis (FMEA) was used to assess the ways an asset or sub-asset can fail and determine the possible effect of that failure.
- SmartSignal alerts were incorporated to leverage physics-based analytics
- Policy designer was used to calculate multiple health indices based on conditional assessment and operational data.
- Rounds was used to gather the conditional assessment and operational data related to the specific assets.
- Queries were used to pull data (e.g. conditional assessments, operating conditions, and technical characteristics) from APM into the policy instances. Queries were also used to cleanse the database of all testing data once the validation was completed
- APM Connect (dataloaders) was used extensively to validate the accuracy of the policy calculations, as well as loading of assets and asset related data such as technical characteristics.
The team from Grid Solutions lent their expertise to two areas of the solution development. The first was development and testing, where the team was trained on APM and was able to quickly translate their decade of maintenance and trouble shouting expertise on T&D assets to create policies, queries, strategies, and more with a minor amount of coaching and guidance. The second was methodology and standardization. In this area, the Grid Solutions team was critical in standardization and calculation of the following health indices:
- AHI – Asset Health Index
- AMI – Asset Maintenance Index
- CPLI – Completeness Index
- POF – Probability of Failure
- ERL – Estimated Remaining Life
- ACI – Asset Criticality Index
- ARI – Asset Risk Index
Effectively modelling and creating a series of blueprints for complex assets such as an static VAR compensator (SVC) was a large task that involved many related and inter-related assets, which effectively formed a performance and reporting hierarchy in addition to the location hierarchy. The performance hierarchy necessitated that the policies be set-up such that they could automatically trigger the re-calculation of health at all levels. An example of this is if something changed at the “PUMP” asset, this would affect the health of the “COOLING” asset which would affect the health of the “SVC” asset.
Reducing time-to-value of electrical grid assets
In the end, the team delivered the Grid APM solution that enables Grid Solutions’ customers to manage their electrical grid assets in such a way that they can capitalize on their maintenance expenses to reduce risks and increase availability. The ready-to-consume content reduces deployment and implementation time/cost by as much as 80%, along with reducing time-to-value by as much as 80%. Grid APM also provides a consistent way to manage all types of assets through common metrics and indicators, as well as the ability to drill down to specific assets and asset attributes. Thanks to GE Digital’s unique set of expertise, applications, and capabilities, Grid Solutions was able to successfully leverage Predix APM to functionally and operationally break the asset down to provide actionable visibility into what matters most. And, this APM technology continues to grow to accommodate more assets, different classes of assets, and different types of data and analytics.