AWS for Industries

Centralized IT/OT Data Management using Element Unify and AWS Industrial Machine Connectivity

To remain competitive, industrial customers need to digitally transform to maximize productivity and asset availability, and to lower costs. To do this, these companies aim to liberate data from legacy operational technology (OT) systems and leverage new tools in the cloud—like machine learning (ML) and artificial intelligence (AI)—to glean new insights from their data. With AWS and our network of leading industrial Partners, you can transform your industrial operations with the most comprehensive and advanced set of cloud solutions available today, while taking advantage of security designed for the most sensitive industries.

AWS for Industrial makes it easy for companies to implement solutions faster using AWS no matter their technology preference. AWS for Industrial offers customers the option to build their own custom solutions using purpose-built industrial building blocks from AWS, work with the AWS Professional Services team and AWS Premier Consulting Partners, or deploy and configure ready-made solutions from AWS Partners for these common use cases.

For example, with large oil and gas customers leveraging multiple systems across the organization, it has become challenging to extract both the real-time and historical OT data along with asset context information to a common indusoperational technology trial data lake for further analysis and inference to improve asset productivity and availability and lower operational costs. These proprietary systems include OSIsoft PI Server (now part of AVEVA), OPC Servers, SCADA platforms, and SAP PM. To this end, as a part of the AWS for Industrial initiative, AWS Industrial Machine Connectivity (IMC) program along with AWS Partner solutions now make it possible to achieve this seamless integration of IT and OT data from multiple systems into an industrial data lake on AWS.

This blog presents the overall end-to-end solution comprising the AWS IMC framework integrated with AWS Partner Element’s Element Unify platform. With this integration, AWS Partners and customers can now centralize, integrate, and contextualize metadata from single-site, multi-site, or multi-instance deployments, and deploy the model(s) to AWS IoT SiteWise. This enables data to be ingested into AWS IoT SiteWise from all Element Unify supported structured and unstructured data sources including but not limited to OPC-UA Server, P&IDs, OSIsoft Asset Framework (AF), and all other supported applications.

AWS Industrial Machine Connectivity

The AWS IMC is an initiative designed to accelerate digital transformation for oil and gas and industrial operators. The AWS IMC Quick Start achieves this by bringing together:

  1. ISV Partners whose edge software products can liberate OT data from on-premises data stores and bring it onto AWS
  2. ISV Partners whose cloud software applications consume that data to generate insights for customers
  3. Regional and Global Systems Integrators (SI/GSIs) that can deliver complete solutions to customers and develop custom applications to meet their unique needs

The IMC framework enables customers and Partners to extract both real-time operational and asset data from industrial assets to AWS in a simple, structured process so that customers can quickly realize the business value that is derived from that data. The framework provides the capability to convert customers’ existing asset hierarchy definitions (such as plants, wells, individual assets) integrated with other industrial information in Element Unify, to the equivalent asset hierarchy within AWS IoT SiteWise.

The following figure is a map of how the IMC kit components relate to ISA 95 levels and what capabilities are enabled at each level. The IMC kit includes AWS certified compatible edge hardware and software components required to connect a customer’s equipment to AWS IoT SiteWise, visualize the data, and run advanced KPI and ML observations using AWS Managed Services. The kit provides automated deployment capability utilizing AWS CloudFormation templates along with all the required documentation. The CloudFormation stack will provision all the relevant AWS Cloud resources for the specific integration and also generate the scripts required to bootstrap physical edge devices. With asset model and real-time SCADA data unified into AWS IoT SiteWise, customers can focus their efforts by building value-added applications and/or connect to other AWS ML/AI and analytics services to extract valuable information that might not have been possible using simple analytics and process trending.

Map showing IMC kit components relation to ISA 95 levels and the capabilities enabled at each level.
Figure1: Map showing IMC kit components relation to ISA 95 levels and the capabilities enabled at each level.

As mentioned earlier, the IMC kit has been designed to enable ISV Partners, such as Element to integrate with its own products and solutions that provide specialized industrial protocol adapters to ingest metadata from different data infrastructure platforms thereby simplifying the process of building IIoT solutions. The source code for the IMC kit is public so Partners and customers will have access to the entire code repository and documentation. SI and GSI Partners are trained to integrate its solution portfolio with the IMC kit, so that the industrial machine connectivity component of the architecture is simplified.

Element Unify

Element Unify allows IT and OT teams to work collaboratively to build rich data context at scale with no-code, automated data pipelines. The platform enables customers to centralize IT/OT metadata management, data model integration, contextualization, and governance for data consumed via AWS IoT SiteWise. AWS Partners and customers can leverage Element Unify to integrate and contextualize IT and OT metadata in single-site, multi-site, multi-instance deployments. Users can also deploy the data model(s) to AWS IoT SiteWise from the customers’ systems (such as PLCs, databases, SCADA systems) through Inductive Automation Ignition and PTC KepServerEx. Element Unify keeps both greenfield and brownfield AWS IoT SiteWise asset models and asset hierarchies updated as the Unify models adapt to changes in the underlying data.

Element Unify closes the gap between raw, siloed, and disorganized IT/OT data and enriched, contextualized, and structured data that can be easily paired with business intelligence (BI), analytical tools, and condition monitoring tools like AWS IoT SiteWise. Element Unify empowers the user to build complex asset data models with ease. The IMC Kit ensures secure and easy provisioning of Element Unify from the AWS Marketplace and enables the ingestion of source system data structures and sensor lists, exported from Ignition and KepServerEx, and stored in Amazon Simple Storage Service (Amazon S3). With the integration with AWS IoT SiteWise, customers can build and deploy data models to AWS IoT SiteWise and keep them up to date with changes happening in source systems. Furthermore, built-in templates and attribute mapping, and other purpose-built transformations and functions make it easy to create the underlying data models to power AWS IoT SiteWise and many other analytical systems including Amazon SageMaker and Amazon QuickSight.

Element Unify is designed to support many real-world use cases such as condition-based monitoring, condition-based and predictive maintenance, and Overall Equipment Effectiveness (OEE) that focus on improving equipment uptime, avoiding revenue loss, cutting Operation and Maintenance costs and improving safety. By connecting together time series, transactional, and static data, it’s easy to develop dashboards for asset health KPIs (such as mean time between failure of asset), process variables trends, maintenance history of assets, asset details (such as make, model), and a hierarchical view of assets within the same dashboard.

The following diagram illustrates the reference architecture for the end-to-end integration between various industrial OT data platforms residing in industrial sites with the Element Unify platform ingesting data into AWS services including AWS IoT SiteWise and Amazon S3 industrial data lake.

Reference Architecture for Element Unify integration with AWS Industrial Machine Connectivity
Figure 2:
Reference Architecture for Element Unify integration with AWS Industrial Machine Connectivity

Now, let’s take a step-by-step view of the reference architecture diagram:

  1. Partner connectivity applications Ignition and KepServerEx running on AWS Certified industrial PC ingests real-time sensor data and metadata from PLC and industrial historians.
  2. PI Interfaces/Connectors collects data from sensors, PLCs, SCADA, into PI Server running at plants or Regional data centers. Element Unify’s AF Management Tool retrieves, deploys, and manages PI Asset Framework asset models, allowing the same asset model to be used in AWS IoT SiteWise. Here the metadata is ingested into Element Unify througha)
    a) File upload on the UI (XLSX, CSV, TSV)
    b) RESTful APIs either directly or via pre-built Unify agents or integration to an iPaaS layer
    c) Directly via command line interface
  3. Element Unify ingests asset models and tag lists exported from edge applications (such as Ignition, Kepware).
  4. Element Unify’s P&ID Productivity Tool automates harvesting of asset information from graphical engineering designs. EAM and ERP systems including SAP FLOC and work orders connected to Element Unify through iPaaS layer.
  5. Element Unify is the centralized IT/OT data management platform that integrates, contextualizes, and governs asset data models from OSIsoft PI Asset Framework (bi-directional), Ignition, Kepware running at the edge and IT data sources like SAP and Maximo, and performs bidirectional model sync with AWS IoT SiteWise.
  6. Element Unify reads existing asset model data from AWS IoT SiteWise and then provisions the new, enriched asset model data back into AWS IoT SiteWise.
  7. AWS IoT SiteWise ingests, filters, transforms, and processes all incoming time series data before storing both raw and processed data in a managed time series data store. AWS IoT SiteWise publishes an MQTT message to AWS IoT Core each time the asset property value updates.
  8. An AWS IoT Core rule publishes these asset property update messages it receives from AWS IoT SiteWise in near real-time into Amazon Kinesis Data Firehose.
  9. Amazon Kinesis Data Firehose captures data from AWS IoT Core, transforms it, and delivers the data in near-real time to Amazon S3 industrial data lake.
  10. Once real-time and historical data is available in Amazon S3 industrial data lake, Amazon Lookout for Equipment can use the data to detect abnormal equipment behavior, so that machine failures can be detected before failure occurs and avoid unplanned downtime. Computed metrics can be written back into Amazon S3, which could then be written back into AWS IoT SiteWise for storage and consumption.
  11. Amazon QuickSight can be leveraged in congestion with Amazon Athena to create and publish interactive BI dashboards.

Conclusion

To remain competitive, industrial customers need to digitally transform to maximize productivity and asset availability, and lower costs. AWS Industrial Machine Connectivity initiative along with Element Unify makes possible for a seamless integration of both real-time and asset context OT data from multiple systems into an industrial data lake on AWS. Now with Element Unify, Asset Health Monitoring can be performed through AWS IoT SiteWise for owner/operators of critical equipment. This enables operations teams to predict and prevent equipment failures using Amazon Lookout for Equipment by providing unified views of time series data from the OSIsoft PI System and other process historians and work order history of assets from EAM systems like SAP PM. Such views become even more valuable with additional analytics, such as, mean time between failure analysis of assets that serve as an indicator for when an asset approaches a failure. Learn more about how industrial customers can optimize operations and digitally transform their business at AWS for Industrial.

Krishna Doddapaneni

Krishna Doddapaneni

Krishna is an IoT Specialist Partner Solutions Architect with AWS, essentially helping partners and customers build crazy and innovative IoT products and solutions on AWS. Krishna has a Ph.D. in Wireless Sensor Networks and a Postdoc in Robotic Sensor Networks. He is passionate about ‘connected’ solutions, technologies, security and their services.

Jie Chou

Jie Chou

Passionate about helping people get the most out of data, Jie Chou leads Integrations and Tools at Element. Prior to the R&D team, he led a team of solution engineers to enable companies to unlock their industrial data. Prior to Element, he was part of the systems engineering team at OSIsoft where he worked with industrial companies on condition-based maintenance and asset health, enabling them to gain more insight from their time-series data. Jie holds a PhD in Bioengineering from Rice University and BS degrees in Electrical Engineering and Bioengineering from University of Illinois at Urbana-Champaign.

Rajesh Gomatam

Rajesh Gomatam

Dr. Rajesh Gomatam is Principal Partner Solutions Architect working for Industrial Software segment at AWS and also leads the AWS Industrial Data Fabric and Computer Vision for Quality Insights solutions. He enjoys working with industrial partners adhering to AWS best practices and has expertise in industrial data platforms, industrial IoT, time series data, analytics, and edge computing. He works closely as a trusted advisor and industry specialist with the partners across manufacturing and energy verticals.

Thomas Cummins

Thomas Cummins

Thomas is an IoT (Internet of Things) Solutions Architect with a passion for developing new software application architectures for IoT gateway devices and sensor networks. Past experience includes IoT systems design and development as well as medical imaging system hardware and software development. As a PhD student in the Biomedical Engineering Department at the University of Southern California, he developed an ultrasound imaging system designed to improve breast cancer diagnosis.